Depth profiling of semiconductor structures using picosecond ultrasound

Abstract

Disclosed herein is a method for depth profiling of a sample including a target region comprising a lateral structural feature. The method includes obtaining a measured signal of the sample and analyzing it to obtain a depth dependence of at least one parameter characterizing the lateral structural feature. The measured signal is obtained by repeating the following: projecting a pump pulse onto the sample, thereby generating an acoustic pulse propagating within the target region; Brillioun scattering a probe pulse off the acoustic pulse within the target region; and detecting a scattered component of the probe pulse to obtain the measured signal. In each repetition, a respective probe pulse is scattered off the acoustic pulse at a respective depth within the target region, thereby probing the target region at multiple depths. A wavelength of the pump pulse is greater than at least about twice a lateral extent of the lateral structural feature.

Description

Technical Field

[0001] This disclosure pertains to a general in-depth analysis of the sample. Background Technology

[0002] Picosecond ultrasound (also known as "picosecond laser ultrasound" and "laser picosecond acoustics") is a non-destructive technique that can be used to obtain structural information from thin films and nanostructures. Typically, an ultrashort optical pulse (often called a "pump pulse") is projected onto the outer surface of the structure. The absorption of the optical pulse heats a thin section (tranche) of the structure adjacent to and including the outer surface. Due to this heating, the section expands, resulting in the formation of an acoustic pulse (also called an "elastic strain pulse" or "strain pulse") that travels deep into the structure and away from the outer surface. Upon reaching a boundary surface (such as the opposite side of a thin film or the second layer of a multilayer structure), at least a portion of the acoustic pulse is reflected and propagates back towards the outer surface. When the acoustic pulse reaches the outer surface, a probe signal is projected onto it for incident on the outer surface. The intensity of the probe signal reflected from the outer surface and the intensity of the reflected component of the probe signal are monitored. From the intensity of the monitored reflected component, one-dimensional structural information about the probed structure can be extracted, such as, for example, film thickness or layer thickness (when the structure is multilayered). Summary of the Invention

[0003] According to some embodiments of this disclosure, aspects of this disclosure relate to depth profiling of samples using picosecond ultrasound. More specifically, but not exclusively, according to some embodiments of this disclosure, aspects of this disclosure relate to picosecond ultrasound-based methods and systems for depth profiling of samples (particularly semiconductor components and structures).

[0004] Therefore, according to aspects of some embodiments, a method for depth profiling of a sample is provided. The method includes the following operations:

[0005] - Provide a sample including the target region. The target region includes lateral structural features.

[0006] - Multiple measured signals can be obtained by performing the following sub-operations multiple times:

[0007] ■ An optical pump pulse is projected onto the sample to generate an acoustic pulse that propagates within a target region of the sample. The wavelength of the pump pulse is at least approximately twice the lateral range of a lateral structural feature along at least one lateral direction.

[0008] ■ An optical probe pulse is projected onto the sample, causing the probe pulse to undergo Brillouin scattering as it leaves the acoustic pulse within the target region.

[0009] ■ Detect the scattered component of the probe pulse to obtain the measured signal.

[0010] In each implementation, the corresponding probe pulse is scattered away from the acoustic pulse at a corresponding depth within the target area, enabling the target area to be detected at multiple depths.

[0011] - Analyze multiple measurement signals to obtain the depth correlation of at least one parameter characterizing the lateral structural features.

[0012] According to some embodiments of the method, the propagation direction of the acoustic pulse within the target region is parallel to the longitudinal dimension of the target region. The longitudinal dimension parameterizes the depth within the sample.

[0013] According to some embodiments of the method, the propagation of acoustic pulses within the target area allows for in-depth profiling of the entire target area.

[0014] According to some embodiments of the method, the probe pulse is configured such that its absorption length in the target region is greater than the range of the target region along the longitudinal dimension.

[0015] According to some embodiments of the method, the wavelength of the pump pulse is at least about twice the lateral range of the lateral structural features along any lateral direction (i.e., all lateral directions).

[0016] According to some embodiments of the method, the lateral structural features manifest as a change in refractive index and / or sound velocity in the target region along at least one lateral direction.

[0017] According to some embodiments of the method, changes in refractive index and / or sound velocity are attributed to one or more changes in a target region along at least one lateral direction, said changes resulting from design (i.e., variations in structure and / or composition specified by the sample design). One or more changes may include the following variations in the target region along at least one lateral direction: geometry, material composition, medium, mass density, density of embedded components and / or pores, and / or spatial arrangement of embedded components and / or pores. At least one parameter characterizing the lateral structural features includes one or more parameters characterizing said geometry, material composition, medium, mass density, density of embedded components and / or pores, and / or spatial arrangement of embedded components and / or pores.

[0018] According to some embodiments of the method, the target region includes one or more embedded components, which include one or more wires, pads, solder bumps, holes, doping concentrations, transistors, transistor components, and / or through-silicon vias.

[0019] According to some embodiments of the method, the pores are formed by the space between parallel fins on the outer surface of the target region.

[0020] According to some embodiments of the method, when analyzing multiple measurement signals, a predetermined desired depth correlation of at least one parameter characterizing the lateral structural features is taken into account.

[0021] According to some embodiments of the method, when analyzing multiple measurement signals, a fitting tool is used to obtain the deep correlation of at least one parameter. The data fitting tool can be derived using machine learning and / or deep learning techniques.

[0022] According to some embodiments of the method, data fitting tools may include regression analysis (e.g., linear regression) and / or artificial neural networks (ANN).

[0023] According to some embodiments of the method, an ANN can be derived using deep learning tools. According to some such embodiments, the ANN can be trained based on data obtained from at least: (i) a previous implementation of the method on a sample manufactured to the same or / or similar design specifications, (ii) a physical modeling of the sample and optionally a system used to implement the method, which may include computer simulations of the sample and the system, and / or (iii) an existing library of measured and / or simulated signals in a similar setting (i.e., similar samples and systems).

[0024] According to some embodiments of this method, the actual depth correlation (geometry and / or composition) of different lateral structural features within different samples can be correlated with multiple measured signals corresponding to the lateral structural features, respectively. The actual (i.e., true) depth correlation of the lateral structural features can be obtained using scanning electron microscopy. The multiple measured signals can be obtained by implementing the signal acquisition operations described above. The obtained correlations can then be used to determine the desired depth correlation of the lateral structural features of the sample to which the method described above is to be performed.

[0025] According to some embodiments of the method, the analysis of multiple measured signals takes into account simulated signals obtained through computer simulation of the sample and the system used to implement the method. The simulated signals include the expected output of a simulated detector under conditions in which lateral structural features and their depth correlations (e.g., equal to design specifications or deviating from design specifications in a defined manner) are pre-specified. The simulated detector models a detector used to obtain multiple measured signals in the operation of detecting scattering components.

[0026] According to some embodiments of the method, the analysis of multiple measured signals includes obtaining the time correlation of the frequency and / or amplitude of the Brillouin oscillations characterizing the scattering component of the probe pulse, and based on this, obtaining the depth correlation of at least one parameter characterizing the transverse structural features.

[0027] According to some embodiments of the method, analyzing multiple measurement signals includes removing the thermo-optical contribution to the multiple measured signals.

[0028] According to some embodiments of the method, pump pulses and / or probe pulses are projected onto the sample so as to be incident on the outer surface of the sample at an incident angle that vanishes (i.e., 0°) or substantially vanishes. The outer surface is parallel to the target region.

[0029] According to some embodiments of the method, the wavelength of the probe pulse is at least about twice the lateral range of the lateral structural feature.

[0030] According to some embodiments of the method, each of the pump pulses is configured to induce mechanical strain in one or more lateral absorption layers of the sample, thereby generating a corresponding acoustic pulse. The one or more absorption layers may be perpendicular to the longitudinal dimension of the target region, which parameterizes the depth within the sample.

[0031] According to some embodiments of the method, each of the pump pulses includes a corresponding pump envelope and a corresponding pump carrier. The pump carrier may be configured to facilitate penetration into the sample and absorption of the pump pulse within the absorption layer. The pump envelope may be configured to facilitate separation of the scattered components from the background signal and noise.

[0032] According to some embodiments of the method, analyzing multiple measured signals includes demodulating each of the measured signals and combining the demodulated signals into a single signal to obtain an extracted signal from the multiple measured signals. According to some embodiments, demodulation can be performed using a lock-in amplifier, with the modulation frequency of the pump pulse (i.e., the shape of the pump envelope) fed as input to the lock-in amplifier.

[0033] According to some embodiments of the method, the absorber layer is silicon-based and the duration of each of the pump pulses is less than 10 psec.

[0034] According to some embodiments of the method, the duration of each of the probe pulses is less than about 10 psec.

[0035] According to some embodiments of the method, the frequency of the pump pulse and / or the frequency of the probe pulse are maximized, or substantially maximized, the intensity of the scattering component of the probe pulse.

[0036] According to some embodiments of the method, the width of the acoustic pulse is less than about 300 nm.

[0037] According to some embodiments of the method, the target region is located within the sample at a distance greater than about 1 μm (micrometers) from the nearest outer surface of the sample, and / or the extent of the target region along the longitudinal dimension is greater than about 2 μm.

[0038] According to some embodiments of the method, each of the pump pulses is configured to induce mechanical strain in the absorber layer by heating the absorber layer(s).

[0039] According to some embodiments of the method, the propagation direction of the acoustic pulse in the target region is perpendicular to (one or more) the absorption layer.

[0040] According to some embodiments of the method, the target region includes one or more absorption layers.

[0041] According to some embodiments of the method, each of the acoustic pulses propagates away from the lateral outer surface of the sample, and the pump pulse and / or probe pulse is projected onto the lateral outer surface.

[0042] According to some embodiments of the method, the absorption layer is positioned within the sample, and each of the acoustic pulses propagates away from the absorption layer toward the lateral outer surface, with the pump pulse and / or probe pulse projected onto said lateral outer surface.

[0043] According to some embodiments of the method, the target region includes multiple lateral structural features. These multiple lateral structural features define a composite lateral structural feature.

[0044] According to some embodiments of the method, the wavelengths of the pump pulse and the probe pulse are configured to allow simultaneous detection of two or more lateral structural features from multiple lateral structural features.

[0045] According to some embodiments of the method, the beam of the probe pulse is equal to or substantially equal to the beam diameter of the pump pulse.

[0046] According to some embodiments of the method, the composite transverse structural features are periodic (i.e., multiple transverse structural features form a repeating transverse structural pattern).

[0047] According to some embodiments of the method, in the operation of analyzing multiple measurement signals, the depth correlation obtained by at least one parameter characterizing the lateral structural features is the average (mean) depth correlation of the multiple lateral structural features.

[0048] According to some embodiments of the method, each of the polarization detection pulses is used to maximize the intensity of the corresponding scattering component.

[0049] According to some embodiments of the method, each of the polarization pump pulses is designed to maximize absorption in the absorption layer, and / or each of the polarization pump pulses is designed to minimize the thickness of the absorption layer, thereby increasing the resolution of the measured signal.

[0050] According to some embodiments of the method, polarization pump pulses and / or probe pulses are used to maximize or substantially maximize the intensity of the scattering component of the probe pulse.

[0051] According to some embodiments of the method, the target region includes a plurality of fins arranged parallel to each other to form a composite lateral structural feature. At least one parameter characterizing the lateral structural feature includes a parameter corresponding to the average width of the fins.

[0052] According to some embodiments of the method, pump pulses and probe pulses are linearly polarized parallel to or substantially parallel to the elongation dimension of the fin.

[0053] According to some embodiments of the method, the elongation dimension of the fin is perpendicular to the longitudinal dimension of the target region, the longitudinal dimension parameterizing the depth within the sample.

[0054] According to some embodiments of the method, the sample is a fin field-effect transistor (FinFET). According to some such embodiments, the width of the acoustic pulse is less than about 50 nm.

[0055] According to some embodiments of the method, the target region includes a plurality of holes projected into the target region along a longitudinal dimension parallel to the target region, the longitudinal dimension parameterizing the depth within the sample. The holes are configured to form a composite lateral structural feature. At least one parameter characterizing the composite lateral structural feature includes a parameter corresponding to the average diameter or average area of ​​the holes.

[0056] According to some embodiments of the method, the holes are arranged in a two-dimensional rectangular array.

[0057] According to some embodiments of the method, a linearly polarized detection pulse is used along a lateral direction parallel to a first direction defined by rows of a two-dimensional rectangular array or a second direction defined by columns of a two-dimensional rectangular array, thereby increasing the measurement sensitivity along the second direction or the first direction, respectively.

[0058] According to some embodiments of the method, the sample is a vertical NAND stack.

[0059] According to some embodiments of the method, the probe pulse is characterized by a first probe wavelength and / or a first probe polarization, and / or the pump pulse is characterized by a first pump wavelength and / or a first pump polarization. The method further includes, prior to the operation of analyzing a plurality of measured signals, repeatedly obtaining a plurality of measured signals by: (i) a second probe pulse characterized by a second probe wavelength and / or a second polarization, and / or (ii) a second pump pulse characterized by a second pump wavelength and / or a second pump polarization, thereby obtaining a second plurality of measured signals. In the operation of analyzing the plurality of measured signals, the plurality of measured signals are analyzed together with at least a second plurality of measured signals.

[0060] According to some embodiments of the method, the operation of obtaining multiple measured signals further includes filtering out, or substantially filtering out, the return component of the pump pulse and / or background and noise before detecting the scattering component of the probe signal.

[0061] According to some embodiments of the method, the sample is a semiconductor component, a part of a semiconductor component, or a semiconductor structure.

[0062] According to some embodiments of the method, the samples are vertical NAND stacks, FinFETs, DRAM structures, gate-surround transistor structures, or phase-change memories.

[0063] According to aspects of some embodiments, a sample analysis system configured to implement the methods described above is provided.

[0064] According to aspects of some embodiments, a computerized system (sample analysis system) for depth profiling of a sample is provided. The system includes an optical setup and a measurement data analysis module. The optical setup is configured to acquire multiple measurement signals from a target region of the sample by repeatedly performing the following operations:

[0065] - An optical pump pulse is projected onto the sample to generate an acoustic pulse that propagates within the target region. The wavelength of the pump pulse is at least approximately twice the lateral range of the lateral structural features in the target region.

[0066] - An optical probe pulse is projected onto the sample, causing the probe pulse to undergo Brillouin scattering as it leaves the acoustic pulse within the target region.

[0067] - Detect the scattering component of the probe pulse, thereby obtaining the corresponding measured signal from multiple measured signals.

[0068] In each repetition, the probe pulse scatters away from the acoustic pulse at a corresponding depth within the target area. The measurement data analysis module is configured to analyze multiple measured signals, or multiple signals derived therefrom (e.g., signals measured using a lock-in amplifier for demodulation), to obtain the depth correlation of at least one parameter characterizing the lateral structural features along at least one lateral direction.

[0069] According to some embodiments of the computerized system, the optical setup includes a signal generating device and a detector. The signal generating device is configured to generate pump pulses and probe pulses and to controllably delay each probe pulse relative to the corresponding pump pulse (thereby allowing control over the depth at which the probe pulses are scattered away from the acoustic pulse). The detector is configured to detect the scattered component of the probe pulses.

[0070] According to some embodiments of the computerized system, the signal generating device includes a variable delay line that allows the delay time of each probe pulse relative to the corresponding pump pulse to be controllably set.

[0071] According to some embodiments of the computerized system, the signal generating device includes at least one light source (e.g., a laser source) configured to generate pump pulses and probe pulses, and at least one optical modulator configured to modulate the pump pulses and / or probe pulses.

[0072] According to some embodiments of the computerized system, the system further includes a lock-in amplifier configured to demodulate the plurality of measured signals before analyzing the plurality of measured signals by a measurement data analysis module.

[0073] According to some embodiments of computerized systems, the signal generating device includes one or more polarization modules that allow controllable setting of the polarization of one or more pump pulses and / or the polarization of one or more probe pulses.

[0074] According to some embodiments of the computerized system, the optical setup further includes a filter (e.g., an optical filter) configured to pass through and transmit scattered components while blocking background signals and noise.

[0075] According to some embodiments of the computerized system, the optical setup further includes an objective lens configured to focus the pump pulse and probe pulse onto the sample.

[0076] According to some embodiments of the computerized system, the optical setup further includes a controller configured to command and coordinate the operation of components of the optical setup (e.g., optical modulators, variable delay lines). The controller may be communicatively associated with a measurement data analysis module.

[0077] According to some embodiments of the computerized system, the optical setup is configured such that the propagation direction of the acoustic pulses within the target region is parallel to the longitudinal dimension of the target region. The longitudinal dimension parameterizes the depth within the sample.

[0078] According to some embodiments of the computerized system, the optical setup is configured such that the propagation of the acoustic pulse within the target area allows for in-depth profiling of the entire target area.

[0079] According to some embodiments of the computerized system, the probe pulse is configured such that its absorption length within the target area is greater than the range of the target area along the longitudinal dimension.

[0080] According to some embodiments of the method, the wavelength of the pump pulse is at least about twice the lateral range of the lateral structural feature along any lateral dimension.

[0081] According to some embodiments of the computerized system, the lateral structural features manifest as changes in the refractive index and / or the speed of sound in a target region along at least one lateral direction.

[0082] According to some embodiments of the computerized system, changes in refractive index and / or sound velocity are attributed to one or more target regions along at least one lateral direction. One or more design changes may include changes to the following along at least one lateral direction of the target region: geometry, material composition, medium, mass density, density of embedded components and / or pores, and spatial arrangement of embedded components and / or pores. At least one parameter characterizing the lateral structural features includes one or more parameters characterizing the geometry, material composition, medium, mass density, density of embedded components and / or pores, and spatial arrangement of embedded components and / or pores.

[0083] According to some embodiments of the computerized system, the target area includes one or more embedded components, which include one or more wires, pads, solder bumps, holes, doping concentrations, transistors, transistor components, and / or through-silicon vias.

[0084] According to some embodiments of the computerized system, the aperture is formed by the space between parallel fins on the outer surface of the target area.

[0085] According to some embodiments of the computerized system, the measurement data analysis module is configured to take into account a predetermined desired depth correlation of at least one parameter characterizing the lateral structural features when analyzing multiple measurement signals.

[0086] According to some embodiments of the computerized system, the measurement data analysis module is configured to employ a fitting tool to obtain the deep correlation of at least one parameter. The data fitting tool can be derived using machine learning and / or deep learning techniques.

[0087] According to some embodiments of computerized systems, data fitting tools may include regression analysis (e.g., linear regression) and / or artificial neural networks (ANNs).

[0088] According to some embodiments of the computerized system, the measurement data analysis module can be configured to take into account simulated signals obtained through computer simulation of the sample and optical setup as characteristics when analyzing multiple measured signals. The simulated signals include the desired output of the detector under conditions in which lateral structural features and their depth correlations (e.g., equal to design specifications or deviating from design specifications in a defined manner) are pre-specified.

[0089] According to some embodiments of the computerized system, the measurement data analysis module may be configured to obtain the time correlation of the frequency and / or the time correlation of the amplitude of the Brillouin oscillation characterizing the scattering component of the probe pulse, and based thereon obtain the depth correlation of at least one parameter characterizing the lateral structural features.

[0090] According to some embodiments of the computerized system, as part of the analysis of multiple measurement signals, the measurement data analysis module can be configured to remove the thermal-optical contribution to the multiple measurement signals.

[0091] According to some embodiments of the computerized system, the optical setup can be configured such that the pump pulse and / or probe pulse are incident on the outer surface of the sample at an incident angle that vanishes (i.e., 0°) or substantially vanishes. The outer surface is parallel to the target region.

[0092] According to some embodiments of the computerized system, the wavelength of the probe pulse is at least about twice the lateral range of the lateral structural feature.

[0093] According to some embodiments of the computerized system, each of the pump pulses is configured to induce mechanical strain in one or more lateral absorption layers of the sample, thereby generating a corresponding acoustic pulse. The one or more absorption layers may be perpendicular to the longitudinal dimension of the target region, which parameterizes the depth within the sample.

[0094] According to some embodiments of the computerized system, each pump pulse includes a corresponding pump envelope and a corresponding pump carrier. The pump carrier can be configured to facilitate penetration into the sample and absorption of the pump pulse within the absorption layer. The pump envelope can be configured to facilitate separation of scattered components from background signals and noise.

[0095] According to some embodiments of the computerized system, the absorber layer is silicon-based and the duration of each of the pump pulses is less than 10 psec.

[0096] According to some embodiments of the computerized system, the duration of each of the probe pulses is less than about 10 psec.

[0097] According to some embodiments of the computerized system, the frequency of the pump pulse and / or the frequency of the probe pulse are maximized, or substantially maximized, the intensity of the scattered component of the probe pulse.

[0098] According to some embodiments of the computerized system, the width of the acoustic pulse is less than about 300 nm.

[0099] According to some embodiments of the computerized system, the target region is located within the sample at a distance greater than about 1 μm (micrometers) from the nearest outer surface of the sample, and / or the range of the target region along the longitudinal dimension is greater than about 2 μm.

[0100] According to some embodiments of the computerized system, each of the pump pulses is configured to induce mechanical strain in the absorber layer by heating the absorber layer(s).

[0101] According to some embodiments of the computerized system, the propagation direction of the acoustic pulse in the target region is perpendicular to (one or more) the absorption layer.

[0102] According to some embodiments of computerized systems, the target region includes one or more absorption layers.

[0103] According to some embodiments of the computerized system, each of the acoustic pulses propagates away from the lateral outer surface of the sample, and the pump pulse and / or probe pulse is projected onto the lateral outer surface.

[0104] According to some embodiments of the computerized system, the absorption layer is positioned within the sample, and each of the acoustic pulses propagates away from the absorption layer toward the lateral outer surface, with the pump pulse and / or probe pulse projected onto the lateral outer surface.

[0105] According to some embodiments of the computerized system, the target region includes multiple lateral structural features. These multiple lateral structural features define a composite lateral structural feature.

[0106] According to some embodiments of the computerized system, the wavelengths of the pump pulse and the probe pulse are configured to allow simultaneous detection of two or more lateral structural features from multiple lateral structural features.

[0107] According to some embodiments of the computerized system, the beam of the probe pulse is equal to or substantially equal to the beam diameter of the pump pulse.

[0108] According to some embodiments of computerized systems, the composite lateral structural features are periodic.

[0109] According to some embodiments of the computerized system, the depth correlation obtained by at least one parameter characterizing the lateral structural features is the average (mean) depth correlation of multiple lateral structural features.

[0110] According to some embodiments of the computerized system, each of the polarization probe pulses is used to maximize the intensity of the corresponding scattering component.

[0111] According to some embodiments of the computerized system, each of the polarization pump pulses is designed to maximize absorption in the absorption layer, and / or each of the polarization pump pulses is designed to minimize the thickness of the absorption layer, thereby increasing the resolution of the measured signal.

[0112] According to some embodiments of the computerized system, polarization pump pulses and / or probe pulses are used to maximize or substantially maximize the intensity of the scattered component of the probe pulse.

[0113] According to some embodiments of the computerized system, the target region includes multiple fins arranged parallel to each other to form a composite lateral structural feature. At least one parameter characterizing the lateral structural feature includes a parameter corresponding to the average width of the fins.

[0114] According to some embodiments of the computerized system, pump pulses and probe pulses are linearly polarized parallel to or substantially parallel to the elongation dimension of the fin.

[0115] According to some embodiments of the computerized system, the elongation dimension of the fin is perpendicular to the longitudinal dimension of the target region, which parameterizes the depth within the sample.

[0116] According to some embodiments of the computerized system, the sample is a fin field-effect transistor (FinFET). According to some such embodiments, the width of the acoustic pulse is less than about 50 nm.

[0117] According to some embodiments of the computerized system, the target region includes a plurality of holes projected into the target region along a longitudinal dimension parallel to the target region, the longitudinal dimension parameterizing the depth within the sample. The holes are configured to form a composite lateral structural feature. At least one parameter characterizing the composite lateral structural feature includes a parameter corresponding to the average diameter or average area of ​​the holes.

[0118] According to some embodiments of the computerized system, the holes are arranged in a two-dimensional rectangular array.

[0119] According to some embodiments of the computerized system, a linearly polarized detection pulse is detected along a lateral direction parallel to a first direction defined by rows of a two-dimensional rectangular array or a second direction defined by columns of a two-dimensional rectangular array, thereby increasing the measurement sensitivity along the second direction or the first direction, respectively.

[0120] According to some embodiments of the computerized system, the sample is a vertical NAND stack.

[0121] According to some embodiments of the computerized system, the probe pulse is characterized by a first probe wavelength and / or a first probe polarization, and / or the pump pulse is characterized by a first pump wavelength and / or a first pump polarization. The optical setup is further configured to acquire a second plurality of measured signals from the target region using a second probe pulse and a second pump pulse. The second probe pulse is characterized by a second probe wavelength and / or a second probe polarization. The second pump pulse is characterized by a first pump wavelength and / or a first pump polarization. The measurement data analysis module is configured to additionally consider the second plurality of measured signals as part of the analysis to obtain the depth correlation of at least one parameter characterizing the lateral structural features.

[0122] According to some embodiments, the sample is a semiconductor component, a part of a semiconductor component, or a semiconductor structure.

[0123] According to some embodiments of the computerized system, the samples are vertical NAND stacks, FinFETs, DRAM structures, gate-surround transistor structures, or phase-change memories.

[0124] According to aspects of some embodiments, a sample analysis system configured to implement the methods described above is provided.

[0125] According to some embodiments, a non-transitory computer-readable storage medium is provided that stores instructions that cause a sample analysis system to perform the methods described above.

[0126] According to aspects of some embodiments, a non-transitory computer-readable storage medium is provided that stores instructions that cause a sample analysis system to perform the following operations:

[0127] - Multiple measured signals of a sample containing a target region, which includes lateral structural features, are obtained by performing the following sub-operations multiple times:

[0128] ■ An optical pump pulse is projected onto the sample to generate an acoustic pulse that propagates within a target region of the sample. The wavelength of the pump pulse is at least approximately twice the maximum lateral dimension of a lateral structural feature along at least one lateral direction.

[0129] ■ An optical probe pulse is projected onto the sample, causing the probe pulse to undergo Brillouin scattering as it leaves the acoustic pulse within the target region.

[0130] ■ Detect the scattered component of the probe pulse to obtain the measured signal.

[0131] In each implementation, the corresponding probe pulse is scattered away from the acoustic pulse at a corresponding depth within the target area, enabling the target area to be detected at multiple depths.

[0132] - Analyze multiple measured signals to obtain the deep correlation of at least one parameter characterizing the lateral structural features.

[0133] According to aspects of some embodiments, a method for depth profiling of a sample is provided. The method includes the following operations:

[0134] - Provide a sample including the target region. The target region includes lateral structural features.

[0135] - Multiple measured signals can be obtained by performing the following sub-operations multiple times:

[0136] ■ An optical pump pulse is projected onto the sample to induce mechanical strain in the sample's transverse absorption layer, thereby generating an acoustic pulse that propagates within the target region of the sample.

[0137] ■ An optical probe pulse is projected onto the sample, causing the probe pulse to undergo Brillouin scattering as it leaves the acoustic pulse within the target region.

[0138] ■ Detect the scattered component of the probe pulse to obtain the measured signal.

[0139] In each implementation, the corresponding probe pulse scatters away from the acoustic pulse at a corresponding depth within the target region, enabling detection of the target region at multiple depths. The absorption layer is positioned within the sample (i.e., such that the surface of the absorption layer does not coincide with the outer surface of the sample), causing each of the acoustic pulses to propagate away from the absorption layer toward the lateral outer surface, onto which the pump pulse and / or probe pulse is projected.

[0140] - Analyze at least a number of measured signals to obtain the deep correlation of at least one parameter characterizing the lateral structural features.

[0141] According to some embodiments of the method, the diameter of the pump pulse beam is approximately equal to or greater than the lateral range of the lateral structural feature.

[0142] According to aspects of some embodiments, a method for depth profiling of a sample is provided. The method includes the following operations:

[0143] - Provide a sample including the target region. The target region includes lateral structural features.

[0144] - An optical pump pulse is projected onto the sample to generate an acoustic pulse that propagates within a target region of the sample. The wavelength of the pump signal is at least approximately twice the lateral range of a lateral structural feature along at least one lateral direction.

[0145] - An optical probe pulse is projected onto the sample, causing the probe signal to undergo Brillouin scattering at multiple depths within the target region as it exits the acoustic pulse.

[0146] - Detect the scattered component of the probe signal to obtain the measured signal.

[0147] - Analyze the measured signals to obtain the deep correlation of at least one parameter characterizing the lateral structural features.

[0148] Some embodiments of this disclosure may include some, all, or none of the advantages described above. One or more other technical advantages will be apparent to those skilled in the art from the accompanying drawings, description, and claims included herein. Furthermore, although specific advantages have been listed above, various embodiments may include all, some, or none of the listed advantages.

[0149] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art. In case of conflict, the patent specification including the definitions shall prevail. As used herein, unless the context clearly indicates otherwise, the indefinite articles “a” and “an” mean “at least one” or “one or more”.

[0150] Unless otherwise specifically stated, as will be apparent from this disclosure, according to some embodiments, terms such as “processing,” “computation,” “operation,” “decision,” “estimate,” “evaluate,” “measure,” etc., may refer to the actions and / or processes of a computer or computing system, or a similar electronic computing component, that manipulates and / or converts data representing physical quantities (e.g., electronic) within the registers and / or memory of the computing system into other data similarly represented as physical quantities within the memory, registers, or other such information storage, transmission, or display devices of the computing system.

[0151] Embodiments of this disclosure may include devices for performing the operations described herein. Devices may be specifically configured for the desired purpose or may include a general-purpose computer selectively enabled or reconfigured by a computer program stored in a computer. Such computer programs may be stored in computer-readable storage media, such as, but not limited to, any type of disk, including floppy disks, optical disks, CD-ROMs, magneto-optical disks, read-only memory (ROM), random access memory (RAM), electrically programmable read-only memory (EPROM), electrically erasable and programmable read-only memory (EEPROM), magnetic or optical cards, or any other type of media suitable for storing electronic instructions and capable of being coupled to a computer system bus.

[0152] The processing and display provided herein are not inherently related to any particular computer or other device. Various general-purpose systems may be used with the teachings and procedures herein, or it may prove convenient to construct more specialized devices to perform the desired methods. The desired architecture of various such systems will be apparent from the description below. Furthermore, embodiments of this disclosure are not described with reference to any particular programming language. It will be understood that various programming languages ​​may be used to implement the teachings of this disclosure as described herein.

[0153] The aspects of this disclosure can be described in the general context of computer-executable instructions (such as program modules) that are executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform a specific task or implement a specific abstract data type. The disclosed embodiments can also be practiced in a distributed computing environment, where tasks are performed by remote processing components connected via a communication network. In a distributed computing environment, program modules may reside in both near-end and remote computer storage media, including memory storage components. Attached Figure Description

[0154] Some embodiments of this disclosure are described herein in conjunction with the accompanying drawings. The description, together with the drawings, enables those skilled in the art to understand how some embodiments can be practiced. The drawings are for illustrative purposes and do not attempt to show structural details of the embodiments in more detail than necessary for a basic understanding of this disclosure. For clarity, some objects depicted in the drawings are not drawn to scale. Furthermore, two different objects in the same drawing may be drawn to different scales. In particular, the scale of some objects may be significantly enlarged compared to other objects in the same drawing.

[0155] In the attached diagram:

[0156] Figures 1A to 1C Three stages in the depth profiling of a sample characterized by lateral structural variations are schematically depicted according to some embodiments.

[0157] Figure 2 A computerized system for depth profiling of samples characterized by lateral structural variations is schematically depicted according to some embodiments.

[0158] Figure 3 A flowchart of a method for depth profiling of a sample characterized by lateral structural variations is provided according to some embodiments.

[0159] Figure 4 A depth profile of a sample characterized by lateral structural variations is schematically depicted according to some embodiments;

[0160] Figure 5 A depth profile of a sample characterized by lateral structural variations is schematically depicted according to some embodiments;

[0161] Figure 6A A schematic perspective view of a sample is provided, which corresponds to a vertical NAND stack according to some embodiments, and can be deeply profiled using the methods and systems of this disclosure according to some embodiments;

[0162] Figure 6B According to some embodiments, Figure 6A A cross-sectional view of the sample;

[0163] Figures 6C to 6F According to some embodiments, schematic depictions are shown respectively in Figure 6A The four stages of in-depth analysis of the sample;

[0164] Figure 6G according to Figures 6C to 6F Some alternative embodiments of those embodiments are illustrated. Figure 6A The stage in the in-depth analysis of the sample;

[0165] Figure 7A A schematic perspective view of a sample corresponding to a FinFET according to some embodiments is presented, and it can be deeply profiled using the methods and systems of this disclosure according to some embodiments thereof;

[0166] Figures 7B to 7E The embodiments are illustratively depicted respectively in Figure 7A The four stages of in-depth analysis of the sample;

[0167] Figure 8 Samples comprising an array of vertical holes are presented according to some embodiments. Figure 3 The simulation results of the implementation of the method are illustrated; the extracted (measured) signal is obtained by scattering the probe pulse away from the acoustic pulse at a depth range within the sample.

[0168] Figures 9A to 9E According to some embodiments, information regarding five samples is provided respectively. Figure 3 Simulation results for five implementations of the method; each sample includes an array of vertical holes; the average depth correlation of the estimated radius of the holes as a function of depth is illustrated.

[0169] Figures 10A to 10E According to some embodiments, respectively, Figures 9A to 9E A cross-sectional view of the sample;

[0170] Figures 11A to 11C According to some embodiments, information regarding three samples is provided respectively. Figure 3 Simulation results for three implementations of the method; each sample includes an array of vertical holes; the average depth correlation of the estimated radius of the holes as a function of depth is illustrated.

[0171] Figure 12A and Figure 12B According to some embodiments, information regarding Figure 8 The sample Figure 3 Simulation results for two implementations of the method are presented; the lateral distribution of the probe pulse intensity is illustrated; the two simulations differ in the polarization of the probe pulse.

[0172] Figure 13 According to some embodiments, samples of this column comprising multiple parallel fins are provided. Figure 3 The simulation results of the method are shown; the extracted (measured) signal is obtained by scattering the probe pulse away from the acoustic pulse at a depth range within the fin.

[0173] Figures 14A to 14E According to some embodiments, information regarding five samples is provided respectively. Figure 3 Simulation results for five implementations of the method; each sample includes multiple parallel fins; the average depth correlation of the estimated fin width as a function of depth is illustrated; and

[0174] Figure 15A and Figure 15B According to some embodiments, information regarding Figure 13 The sample Figure 3 Simulation results for two implementations of the method are presented; the distribution of the probe pulse intensity is illustrated; the two simulations differ in the polarization of the probe pulse and the pump pulse. Detailed Implementation

[0175] The principles, uses, and implementations of the teachings herein can be better understood by referring to the accompanying description and figures. After carefully reading the description and figures provided herein, those skilled in the art will be able to implement the teachings herein without excessive effort or experimentation. In the figures, the same reference numerals refer to the same parts throughout the text.

[0176] In the specification and claims of this application, the terms “comprising” and “having”, and their forms are not limited to members of the list that the terms may be associated with.

[0177] As used herein, the term "substantially" can be used to specify that a first property, quantity, or parameter is close to or equal to a second property, quantity, or parameter, or a target property, quantity, or parameter. For example, when the length of a first object is measured to be at least 80% (or some other predefined threshold percentage) and no more than 120% (or some other predefined threshold percentage) of the length of a second object, the first object and the second object can be said to have "substantially the same length." In particular, the case where the first object has the same length as the second object is also covered by the expression that the first object and the second object have "substantially the same length."

[0178] According to some embodiments, the target quantity may refer to an optimal parameter, which can in principle be obtained using mathematical optimization software. Thus, for example, when the parameter value is equal to at least 80% of the maximum possible value (or some other predefined threshold percentage), the value assumed by the parameter may be described as "substantially equal to" the maximum possible value that can be assumed by the parameter. In particular, the case where the parameter value is equal to the maximum possible value is also covered by the statement that the value assumed by the parameter is "substantially equal to" the maximum possible value that can be assumed by the parameter.

[0179] As used herein, the term "about" can be used to specify the value of a quantity or parameter (e.g., the length of a component) as a range of values ​​near (and including) a given (stated) value. According to some embodiments, "about" can specify a parameter value between 80% and 120% of a given value. For example, stating "the length of the component is about 1 m" is equivalent to stating "the length of the component is between 0.8 m and 1.2 m." According to some embodiments, "about" can specify a parameter value between 90% and 110% of a given value. According to some embodiments, "about" can specify a parameter value between 95% and 105% of a given value.

[0180] As used herein, the terms “substantially” and “about” are used interchangeably according to some embodiments.

[0181] For ease of description, a three-dimensional Cartesian coordinate system (with orthogonal axes x, y, and z) is introduced in some of the accompanying figures. Note that the orientation of the coordinate system relative to the depicted object can vary between the figures. Additionally, the symbol ⊙ can be used to indicate axes pointing "outside the page," while the symbol... It can be used to represent an axis pointing "within the page".

[0182] Referring to the accompanying drawings, in the block diagram and flowchart, optional components and operations can appear within boxes drawn by dashed lines.

[0183] This disclosure advantageously extends picosecond ultrasound technology to allow three-dimensional probing of samples such as semiconductor components, characterized by one or more (physical) properties that can vary along one or more lateral directions. The manner in which these properties vary can be depth-dependent. Therefore, this disclosure summarizes picosecond ultrasound technology to allow obtaining not only one-dimensional structural information about the probed structure, but also two-dimensional and three-dimensional structural information. In particular, the methods and systems of this disclosure advantageously allow three-dimensional probing of nanostructures within samples (e.g., wafers) that are located too deep within the sample or extend too far into the sample, allowing for probing using scanning electron microscopy.

[0184] More specifically, this disclosure extends picosecond ultrasound technology to allow for the deep profiling of one or more “lateral structural features.” That is, the geometry, density, and / or material composition of the structural feature varies, for example, along at least one lateral dimension, and its lateral extent can be as small as 5 nm (at least along one lateral direction). In particular, the methods and systems of this disclosure enable the estimation of the depth correlation of one or more parameters of a parameterized lateral structural feature (e.g., one or more parameters characterizing variations in geometry and / or density along one or more lateral directions).

[0185] This is a non-restrictive example intended to make the explanation more specific. Figures 1A to 1C According to some embodiments of this disclosure, different stages in the depth profiling of a sample 100 (e.g., a semiconductor component) characterized by lateral structural variations are schematically depicted in the direction including lateral structural features. More precisely, a target region 110 of the sample 100 is depicted. Dashed line B1 indicates a first (lower, as depicted in the figures) boundary of the target region 110, and dashed line B1' indicates a second (upper, as depicted in the figures) boundary of the target region 110. The target region 110 includes three adjacent sub-regions: a first side sub-region 110a, a second side sub-region 110b, and an intermediate sub-region 110c. The intermediate sub-region 110c is located between the first side sub-region 110a and the second side sub-region 110b. The side sub-regions 110a and 110b may, for example, correspond to a first solid medium characterized by a first refractive index, while the intermediate sub-region 110c may, for example, correspond to a second solid medium characterized by a second refractive index different from the first refractive index.

[0186] Additionally or alternatively, according to some embodiments, the first solid medium and the second solid medium may be characterized by different sound velocities. In this regard, it is noted that, in order to achieve the depth profiling method disclosed herein, variation of at least one of the refractive index and sound velocity across the target region is sufficient.

[0187] According to not in Figures 1A to 1C In some alternative embodiments depicted, the intermediate sub-region 110c may form a pore.

[0188] although Figures 1A to 1C A two-dimensional view of the target region 110, intercepted along the zx plane, is provided for description purposes, assuming that the target region 110 is uniform along the y-axis until the manufacturing defects are uniform. Therefore, Figures 1A to 1C The target area 110 was effectively and completely depicted.

[0189] exist Figures 1A to 1CIn the target region 110, the width of the intermediate sub-region 110c decreases as the z-coordinate increases. Therefore, the structure of the target region 110 exhibits not only lateral (i.e., transverse) variations (or correlations) but also longitudinal variations. In other words, the geometry of the target region 110 varies not only along the x-axis but also along the z-axis (for some range of x values). More specifically, as the target region 110 traverses parallel to the x-axis from its first boundary indicated by dashed line B1 to its second boundary indicated by dashed line B1', its material composition changes twice. In this sense, the structure of the target region 110 can be described as including lateral structural features. More specifically, as the target region 110 traverses parallel to the x-axis, the lateral structural features consist of changes from a first medium (characterized by a first refractive index) to a second medium (characterized by a second refractive index) and back to the first medium.

[0190] The shape of the intermediate sub-region 110c can be estimated using the methods and systems disclosed herein. More precisely, the depth correlation (i.e., correlation with the z-coordinate) of the width of the intermediate sub-region 110c (indicated by w(z)) can be estimated using the disclosed methods and systems. In this sense, the disclosed methods and systems are referred to as allowing depth profiling of samples (e.g., semiconductor components). In particular, this disclosure teaches how to estimate the depth correlation of parameters characterizing lateral structural features. Figures 1A to 1C In this context, the natural choice of parameter is the width of the intermediate subregion 110c (indicated by w(z)). Therefore, the function w(z) can be evaluated using the method and system of this disclosure.

[0191] The target region 110 further includes a lateral absorption layer 120. According to some embodiments, and as in... Figures 1A to 1C As depicted, the absorber layer 120 is positioned adjacent to the outer surface 124 of the sample 100. See also Figure 1A A pump pulse 123 (e.g., a laser pulse) is projected onto a region of the outer surface 124 adjacent to the absorption layer 120. The pump pulse 123 is configured to be absorbed in the absorption layer 120, thereby heating the absorption layer 120. Note that the (lateral) area of ​​the absorption layer 120 depends on the beam width DIA1 (i.e., beam diameter; indicated by the double-headed arrow) of the pump pulse 123. The increase in temperature (of the absorption layer 120) causes mechanical strain in the absorption layer 120, resulting in the generation of an acoustic pulse 125. Figure 1B and Figure 1C (As shown). The acoustic pulse 125 propagates away from the outer surface 124 and into the depth of the target region 110.

[0192] As acoustic pulse 125 propagates within target region 110, it temporarily and locally modifies the density distribution of the segment instantaneously located across the acoustic pulse. This, in turn, leads to a temporary modification of the refractive index attributable to elastic optical effects. These changes in refractive index can be sensed by Brillouin scattering. More specifically, and as detailed below, a series of probe pulses can each scatter away from a series of acoustic pulses (such as acoustic pulse 125) at a corresponding depth, and the corresponding (backward) scattered components of the probe pulses can be detected to obtain multiple measured signals. The multiple measured signals can be analyzed to reveal structural information about the probed structure, such as, for example, the width w(z) of the intermediate sub-region 110c.

[0193] The beam diameter of pump pulse 123 may be greater than or at least equal to the maximum width of the intermediate sub-region 110c, or in mathematical terms, DIA1 ≥ w. max , where w max =max z [w(z)] (i.e., the maximum value that can be assumed from w(z)). This further implies that the lateral range of the acoustic pulse 125 is at least approximately equal to the maximum width of the intermediate sub-region 110c, in which case it can constitute the lateral range of the lateral structural features (i.e., the maximum lateral range). Considering optical diffraction limitations, in order to produce a well-defined beam, the wavelength of the pump pulse 123 cannot be less than about 2·DIA1. Therefore, the wavelength of the pump pulse 123 can be greater than or approximately equal to 2·w max .

[0194] Despite Figures 1A to 1C In the text, the width of the target region 110 is depicted as greater than the maximum width of the intermediate sub-region 110c (i.e., w). max It will be understood that the width of the target region 110 can be selected to be narrower. Specifically, the width of the target region 110 can be selected to be equal to w. max (By selecting pump pulse 123, the beam diameter is equal to w) max ).

[0195] See Figure 1C A probe pulse 127 (e.g., a laser pulse) is projected onto a region of the outer surface 124 adjacent to the absorption layer 120. The probe pulse 127 is configured to penetrate into the target region 110 so as to undergo Brillouin scattering away from the acoustic pulse 125 (within the target region 110). A scattering component 131 of the probe pulse 127 that is backscattered away from the acoustic pulse 125 is further indicated. The scattering component 131 can be detected by a (optical) detector 132 to generate a corresponding measured signal.

[0196] The depth of the target region (or the equivalent z-coordinate) is determined by the depth at which the probe pulse 127 "intercepts" the acoustic pulse 125 (i.e., experiences scattering from it). The z-coordinate at which the probe occurs can be selectively controlled by adjusting the time delay Δt between the emission of the pump pulse 123 and the emission of the probe pulse 127. By implementing the operation described above with different time delays Δt, the target region 110 can be probed at multiple corresponding depths, and in particular, always along the depth dimension of the target region 110. In this way, multiple measured signals corresponding to multiple (probe) depths can be obtained.

[0197] As described in detail in the Methods section below, multiple measured signals can be analyzed to evaluate (estimate) w(z), or in other words, to estimate the correlation between the width and depth of the intermediate sub-region 110c.

[0198] Note that the measurement resolution along the z-axis is determined by the width u of the acoustic pulse 125, which in turn can be determined by the thickness b of the absorbing layer 120. Figure 1B and Figure 1C As indicated in the instructions below, which are discussed in more detail in the Methods section, u can be advantageously as small as about 10 nm, according to some embodiments.

[0199] Finally, it should be noted that, although in Figures 1A to 1C In the illustration, the absorber layer 120 is shown as the outer surface 124 adjacent to the sample 100; other options are generally possible, for example, as in Figure 4 and Figure 5 As depicted in the text. In particular, according to some embodiments, the absorbent layer may be completely embedded within the sample region and / or positioned outside the target region.

[0200] As used herein, according to some embodiments, the term "lateral structural feature" refers to a structure that varies laterally along at least one (lateral) direction, in a sense that the value of at least one parameter characterizing the structure is not constant along that direction. A lateral structural feature can manifest as a change in one or more of the following: geometry, material composition, medium, mass density, density of embedded components and / or pores, spatial arrangement of embedded components and / or pores, doping concentration (i.e., density of doped impurities), which in turn manifests as a change in refractive index and / or sound velocity along at least one lateral direction. Changes in optical properties along at least one lateral direction, such as, for example, changes in birefringence and / or optical anisotropy (i.e., the correlation between refractive index and polarization and / or the direction of light propagation therein), can also constitute a lateral structural feature in the sense described above.

[0201] Note that lateral structural features can vary along two lateral directions or only along a single lateral direction. The disclosed methods and systems allow for in-depth profiling in both cases. Figure 4 and Figures 6A to 6G Examples of lateral structural features that vary laterally along two (lateral) directions are depicted. Figures 1A to 1C , Figure 5 ,and Figures 7A to 7E An example of a lateral structural feature (up to a manufacturing defect) that varies laterally only along a single direction (i.e., parallel to the x-axis) is depicted.

[0202] Generally, the target region is selected to completely encompass the lateral structural features. If the lateral structural features form portions of larger features extending beyond the target region, no changes in the larger features outside the target region are detected (and variations along one or two lateral directions do not affect the classification of the lateral structural features). Furthermore, the size of the target region can be increased (by increasing the beam width of the pump pulse and probe pulse) to completely encompass the larger features.

[0203] As used herein, according to some embodiments, the term "lateral extent," referring to a lateral structural feature that varies laterally along two lateral directions, can refer to the maximum lateral extent of the feature (where the maximum value is determined in the depth dimension). For example, the lateral extent of a (circular) cylinder whose axis of symmetry is parallel to the longitudinal direction is the diameter of the cylinder, while the lateral extent of a truncated cone whose axis of symmetry is parallel to the longitudinal direction is the larger of the two diameters. The lateral extent of an elliptical cylinder whose axis of symmetry is parallel to the longitudinal direction (i.e., a cylinder whose lateral cross-section defines an ellipse) is the larger of the two diameters of the ellipse.

[0204] According to some embodiments, the term "lateral range" with reference to a lateral structural feature that varies laterally only along a single lateral direction may refer to the maximum range of the feature along that lateral direction (where the maximum value is determined in the depth dimension). According to some embodiments, the term "lateral range" with reference to a lateral structural feature whose rate of change along a first lateral direction is significantly greater than that along a second lateral direction may refer to the maximum range of the feature along the first lateral direction.

[0205] According to some embodiments, the term "lateral range," referring to a lateral structural feature that varies laterally along two lateral directions, may refer to the maximum extent of the feature along a particular lateral direction, regardless of whether the maximum extent of the feature along the other lateral direction is large. This may be the case, for example, in embodiments where an attempt is made to analyze the lateral structural feature in depth along only one lateral direction.

[0206] system

[0207] Figure 2A computerized system 200 for depth profiling of samples, such as semiconductor components and structures, is schematically depicted according to some embodiments. System 200 includes a light source 202, a detector 204 (optical sensor), a measurement data analysis module 208, a platform 212, a controller 214, optical devices 216, and a lock-in amplifier 218. The light source 202 may be a coherent light source, such as a laser source (i.e., a laser generator). Optical devices 216 may include a pump modulator 222, a variable delay line 226, a filter 230 (e.g., an optical filter), and optionally one or more of the following: a probe modulator 232, a pump polarization module 236, and / or a probe polarization module 238. Optical devices 216 may further include an objective lens 244 and multiple beam splitters 246: a first beam splitter 246a, a second beam splitter 246b, and a third beam splitter 246c.

[0208] The light source 202, detector 204, platform 212, and optical device 216 constitute the optical setup (unnumbered) of system 200 or a part of the optical setup.

[0209] Platform 212 is configured to place a sample, such as sample 250 (e.g., a semiconductor component or structure), thereon. A target region 252 in sample 250 to be probed (i.e., in-depth profiling) is also indicated. As detailed below, target region 252 may include lateral structural features (not shown in the original text). Figure 2 (See illustration). That is, a structure that varies along at least one lateral direction (i.e., parallel to the xy plane) means that the value of at least one parameter characterizing the structure is not constant along that direction. According to some embodiments, the structure of the target region 252 may further vary along a longitudinal direction (i.e., along the z-axis, its quantization depth). According to some embodiments, the target region 252 may include multiple lateral structural features constituting a composite lateral structural feature (i.e., multiple identical lateral structural features). According to some such embodiments, the multiple lateral structural features constitute a repeating lateral structural pattern (i.e., a periodic composite lateral structural feature).

[0210] According to some embodiments, the target region 252 may include multiple different lateral structural features. According to some embodiments, particularly when not periodically arranged, the multiple lateral structural features may be collectively regarded as a single lateral structural feature.

[0211] Controller 214 may be functionally associated with each of the following: light source 202, detector 204, platform 212, optical device 216, and lock-in amplifier 218. More specifically, controller 214 is configured to control and synchronize the operation and function of the modules and components listed above, particularly those of optical device 216, during depth profiling of the sample. For example, controller 214 may set a series of time delays imparted by variable delay line 226, such that a minimum time delay allows detection of target region 252 at maximum depth, and a maximum time delay allows detection of target region 252 at minimum depth.

[0212] According to some embodiments, platform 212 is movable at least along one or more lateral directions, thereby allowing for in-depth profiling of different target regions in the sample. According to some embodiments, platform 212 may be configured to allow monitoring and control of the temperature of the sample placed thereon. For example, according to some embodiments, the sample placement surface of platform 212 (i.e., the top surface of platform 212) may be controllably cooled (and optionally heated).

[0213] In operation, the light source 202 may generate a laser beam 215 (e.g., a laser pulse) guided at the first beam splitter 246a. According to some embodiments, the laser beam 215 may be a laser pulse or may comprise a series of pulses. A first sub-beam 215a (also referred to as a probe pulse 227) indicates the portion of the laser beam 215 that passes through the first beam splitter 246a. A second sub-beam 215b indicates the portion of the laser beam 215 reflected from the first beam splitter 246a toward the pump modulator 222. The second sub-beam 215b is modulated by the pump modulator 222, thereby preparing a pump pulse 223 (indicated by the dashed arrow), as detailed below. The pump pulse 223 travels from the pump modulator 22 toward the objective lens 244 (via the second beam splitter 246b) and is thus focused onto the sample 250. The pump pulse 223 is configured to be absorbed by the absorption layer of the sample 250 (not in…). Figure 2 (as shown in the diagram) absorption and thus generation of sound pulses (not shown in the diagram) Figure 2 (As shown in the diagram), basically as Figures 1A to 1C The description is as follows and will be further elaborated below. Some possible configurations of the target region and the absorption layer in the sample are described in... Figures 1A to 1C , Figure 4 ,and Figure 5 In China, and in Figures 6A to 6G and Figures 7A to 7E Described in the text.

[0214] According to some embodiments, pump modulator 222 may be configured to modulate the waveform of the second sub-beam 215b such that pump pulse 223 is characterized by pump carrier (i.e., carrier wave) and pump envelope: pump carrier may be configured (e.g., by wavelength characterization) to facilitate the penetration of pump pulse 223 into the sample, and when the absorption layer is fully embedded in the sample (e.g., as in...). Figure 4 and Figure 5 (As depicted in the diagram), the pump envelope propagates onto the absorption layer and absorbs the pump pulse 223 within the absorption layer, as described below in the Methods section. The pump envelope can be configured to facilitate the separation of the scattered components of the probe pulse from background signals and noise, thus improving detection. According to some such embodiments, the pump modulator 222 may include a frequency multiplier (not shown).

[0215] According to some embodiments, a portion of the pump pulse 223 may return from the sample 250 (due to one or more scattering and / or reflection mechanisms). This return component 251 of the pump pulse 223 is... Figure 2 The dotted double-pointed arrow indicates this. As explained in detail below, return component 251 can be largely filtered out by filter 230.

[0216] According to some embodiments, the optical device 216 further includes a pump polarization module 236 (e.g., a polarization filter with controllable polarization angle), which can be used to modify the polarization of the pump pulse 223 (e.g., from circular polarization to linear polarization) to maximize or substantially maximize the absorption of the pump pulse 223 within the absorption layer (e.g., as in...). Figure 7A and Figure 7B (as described in the description), and thereby increases the amplitude of the Brillouin oscillation, thus potentially facilitating its extraction by the measurement data analysis module 208.

[0217] The probe pulse 227 (i.e., the first sub-beam 215a) travels from the first beam splitter 246a to the variable delay line 226. As described in detail below, the variable delay line 226 is configured to delay the probe pulse 227 for a controllably selectable time interval. According to some embodiments, the delayed probe pulse 227 (indicated by the dashed arrow) continues from the variable delay line 226 to the third beam splitter 246c. According to some embodiments, and as in... Figure 2 As depicted, on the path to the third beam splitter 246c, the probe pulse 227 can pass through the probe modulator 232. In this embodiment, the probe pulse 227 can be modulated by the probe modulator 232 according to a modulation signal received from the controller 214. According to some embodiments, and as shown in Figure 2 As depicted, the probe pulse 227 can be further passed through the probe polarization module 238 (e.g., a polarization filter whose polarization angle can be controllably selected).

[0218] The probe pulse 227 is reflected from the third beam splitter 246c toward the second beam splitter 246b. The probe pulse 227 is then reflected from the second beam splitter 246b toward the objective lens 244, which focuses the probe pulse 227 onto the sample 250. As described in detail below, the probe pulse 227 is configured to penetrate the sample 250 and enter the target region 252 so as to scatter away from the acoustic pulse at a depth within the target region 252, which is determined by the delay time imparted by the variable delay line 226. More specifically, the variable delay line 226 can be configured to controllably increase the optical path length of the probe pulse 227 (e.g., using a mirror), thereby increasing its travel time, resulting in the probe pulse 227 arriving at the sample 250 with a controllable time delay relative to the pump pulse 223. A (backward) scattering component 231 of the probe pulse 227 is also indicated (because it scatters away from the acoustic pulse).

[0219] According to some embodiments, the optical device 216 further includes a detection polarization module 238, which can be used to modify the polarization of the detection pulse 227 in order to maximize or substantially maximize the intensity of the scattering component 231 of the detection pulse 227.

[0220] Filter 230 may be configured to allow its transmitted scattering component 231 to pass through while filtering out noise and / or blocking the return component 251. According to some embodiments, filter 230 is or includes an optical filter, and the wavelengths of the pump pulse 223 and the probe pulse 227 may be selected such that the wavelengths characterizing the return component 251 and the scattering component 232 are different, thereby allowing the scattering component 231 to transmit through filter 230 and thereby blocking the return component 251. More generally, according to some embodiments, the waveforms of the pump pulse 223 and the probe pulse 227 are selected to allow for differentiation therebetween using an optical filter. According to some embodiments, the rays of the scattering component 231 and the return component 251 arriving at filter 230 may be oriented along two different directions (or two different angular ranges). In such embodiments, filter 230 may be or also include an angle filter (i.e., allowing only light arriving at a specific angle of incidence to transmit through).

[0221] According to some embodiments including a probe polarization module 238 and an optional pump polarization module 236, the filter 230 may be or include a polarization filter. In such embodiments, the polarization of the probe pulse 227 and optionally the pump pulse 223 may be selected, for example, to allow the scattered component 231 to pass through the filter 230 and to block or substantially block the return component 251.

[0222] Detector 204 is configured to detect the output of filter 230 (i.e., the scattered component 231 after it has passed through filter 230) to obtain a measured signal. The measured signal can be relayed to lock-in amplifier 218. Lock-in amplifier 218 is configured to receive from controller 214 a modulated signal employed by pump modulator 222 in preparation of pump pulse 223. Lock-in amplifier 218 uses the modulated signal to obtain an extracted (i.e., FM demodulated) signal, where the contribution of the scattered component 231 to the measured signal is amplified (and background signal and noise are suppressed). The extracted signal essentially represents the deviation from the baseline signal (which would otherwise be obtained in the absence of a pump pulse). Therefore, the extracted signal corresponds to the Brillouin oscillation attributed to the probe pulse scattering away from the acoustic pulse (which is in turn generated by the pump pulse).

[0223] To estimate the depth correlation of lateral structural features (e.g., to estimate w(z) in a specific embodiment of sample 250, where sample 250 has the structure of sample 100), the detection operation of the sequence described above can be implemented at multiple different time delays to detect the target region at corresponding multiple different depths. For each time delay, the corresponding measured signal M is obtained. r (z). M r (z) indicates the signal measured when the sound pulse 227 is scattered away from the acoustic pulse at coordinate z (within the target region 252). Multiple measured signals {M r (z)} r (For example, after demodulation) the measurement data analysis module 208 can analyze the data to obtain an estimate of the depth correlation of the lateral structural features, as detailed below. (Note that the same depth can be detected multiple times.) The number of signals measured and the corresponding time delay may depend on the longitudinal extension of the target area. As a non-limiting example, according to some embodiments, the time delay may be less than about 2 nsec (nanoseconds), where the time sampling resolution is about 1 psec (picosecond).

[0224] Measurement data analysis module 208 includes computer hardware (one or more processors, and volatile and non-volatile memory components; not shown). Measurement data analysis module 208 can be configured to receive (e.g., one at a time) data corresponding to {M} from lock-in amplifier 218. r (z)} r The extracted signals are combined into a (single) combined signal ES(z). The measurement data analysis module 208 is further configured to analyze the combined signal to obtain the depth correlation (or equivalently, the correlation with the z coordinate) of at least one parameter characterizing the lateral structural features within the target region 252.

[0225] Additionally, according to some embodiments, the measurement data analysis module 208 may be configured to isolate the elastic optical contribution to the measured signal from the contributions to the measured signal attributable to other physical effects triggered by the pump pulse. Specifically, as detailed below in the Methods section, based on their different physical characteristics, the measurement data analysis module 208 may be configured to distinguish between the elastic optical contribution to the measured signal and the thermo-optical contribution to the measured signal. Using signal processing techniques, the thermo-optical contribution to the (extracted) measured signal can be identified and subtracted from it.

[0226] According to some embodiments, the target region 252 includes a composite lateral structural feature (not shown) composed of multiple lateral structural features that are identical up to the manufacturing defect. Additionally, the beam diameter of the probe pulse 227 is selected to allow complete probe of the target region 252. In this embodiment, each of the multiple lateral structural features contributes substantially equally to the combined signal. The output of the measurement data analysis module 208 can then be interpreted as the average depth correlation characterizing at least one parameter of the lateral structural features (i.e., averaging over all lateral structural features included in the multiple lateral structural features).

[0227] According to some embodiments, the measurement data analysis module 208 can be configured to obtain the depth correlation of at least one parameter characterizing lateral structural features based on two or more measured signals. Each of the two or more measured signals can be obtained accordingly for a unique pump pulse-probe pulse combination. According to some embodiments, different pump pulse-probe pulse combinations can vary from one to another among one or more optical characteristics selected from: the wavelength of the pump pulse, the waveform of the pump pulse, the polarization of the pump pulse, the wavelength of the probe pulse, the waveform of the probe pulse, and the polarization of the probe pulse.

[0228] According to some embodiments, where the sample is a wafer, system 200 can be configured to perform the sequence of probing operations described above at different locations on the wafer. Measurement data analysis module 208 can be configured to analyze multiple measured signals to obtain information about processing variations across the wafer.

[0229] According to some embodiments, the same location(s) on different dies(s) can be detected and the multiple measured signals obtained can be subsequently compared to obtain a large-scale map of the processing variation across the wafer (e.g., as part of a variation protocol between dies).

[0230] According to some embodiments, different locations on the same die, known to be characterized by the same lateral structural features (such as different regions in a memory array sharing the same architecture) up to manufacturing defects, can be detected, and multiple measured signals can then be compared to obtain a small-scale map of processing variations across the die or one or more regions thereof (e.g., as part of a variation protocol within the die). In this regard, it is noted that the type of deviation from the design specifications of the lateral structural features(s) can depend on the density of the lateral structural features in the regions on the die (e.g., when the lateral structural features constitute a repeating lateral structural pattern).

[0231] The various ways in which the measurement data analysis module 208 can process measurement signals to obtain the depth correlation of one or more parameters characterizing the lateral structural features are further described in the Methods section below.

[0232] According to some embodiments, the optical device 216 may be configured such that each of the pump pulse 223 and the probe pulse 227 is incident on the sample 250 at a vanishing or substantially vanishing angle of incidence (i.e., an angle of incidence equal to or substantially equal to zero).

[0233] According to some embodiments, the optical device 216 may include optical components configured to allow controllable modification of the angle of incidence of the pump pulse 223 and / or the probe pulse 227.

[0234] According to not in Figure 2 In some embodiments depicted, system 200 may include two light sources: a first light source configured to generate pump pulse 223 and a second light source configured to generate probe pulse 227.

[0235] method

[0236] Figure 3 This is a flowchart of a method 300 for depth profiling of a sample including one or more lateral structural features, according to some embodiments. Method 300 may be implemented by a computerized system 200 or a similar computerized system. Method 300 may include the following operations:

[0237] - Operation 310, in which a sample is provided. The sample includes a target region that includes lateral structural features.

[0238] - Operation 320, in which multiple measured signals are obtained by performing the following items m times:

[0239] ■ Sub-operation 320a, wherein an optical pump pulse is projected onto the sample to generate an acoustic pulse that propagates within a target region. The wavelength of the pump pulse is selected to be at least approximately twice the lateral range of the lateral structural features.

[0240] ■ Sub-operation 320b, in which an optical probe pulse is projected onto the sample such that the probe pulse undergoes Brillouin scattering from an acoustic pulse within the target region.

[0241] ■ Sub-operation 320c, in which a scattered component of the probe pulse is detected to obtain a measured signal.

[0242] In each of the m realizations, the corresponding probe pulse is scattered from the acoustic pulse at a corresponding depth within the target region such that the target region is probed at multiple depths. (For example, according to some embodiments, in the i-th and j-th realizations (i < j ≤ m), the i-th probe pulse may be scattered at the i-th depth s i and the j-th probe pulse may be scattered at the j-th depth (s j ≠ s i ).)

[0243] - Operation 330, in which the depth dependence of at least one parameter characterizing a lateral structural feature is obtained by analyzing at least a plurality of measured signals (obtained in the m realizations of sub-operations 320a - 320c).

[0244] According to some embodiments, the time delays of the corresponding probe pulses relative to the corresponding pump pulses for different realizations (from the m realizations) may be different from each other, while the set parameters (e.g., wavelengths of the pump and probe pulses, their polarizations, etc.) may be the same in each realization. According to some embodiments, the time delay may measure the time interval between the incidence of the pump pulse on the sample and the incidence of the probe pulse on the sample. More specifically, according to some embodiments, the m time delays may be selected such that the probe pulses are scattered from the acoustic pulses at multiple depths within the target region: for example, the first probe pulse (or the first set of probe pulses, e.g., when each depth is probed more than once) is scattered from the first acoustic pulse at the first depth s1, the second probe pulse (or the second set of probe pulses) is scattered from the second acoustic pulse at the second depth s2, and so on, until the m-th probe pulse (or the m-th set of probe pulses) is scattered from the m-th acoustic pulse at the m-th depth s m and s1 > s2 >... > s m .

[0245] In sub-operation 320a, the wavelength of the pump pulse may be selected to maximize or substantially maximize the absorption of the pump pulse in an absorption layer, such as absorption layer 120 or in Figure 4 , Figure 5 , Figures 6D to 6G , and Figures 7C to 7EThe absorption layer is depicted in the image. More precisely, the absorption layer can be understood as a layer in the sample in which most of the pump pulses are absorbed. The thickness of the absorption layer can depend on the absorption length of the pump pulses in the medium or medium within the region of the sample in which it constitutes the sample (where the absorption layer is located). Therefore, by increasing the absorptivity of the pump pulses in the region (where the absorption layer is located) (e.g., by suitably changing their wavelength), the absorption length of the pump pulses in the region and thus the thickness of the absorption layer decrease. Since the width of the acoustic pulse (which determines the resolution) can depend on the thickness of the absorption layer, this can in turn lead to an increase in the resolution of the measured signal.

[0246] The last statement remains true as long as the duration of the pump pulse is shorter than the formation time of the acoustic pulse (i.e., the thermal expansion time of the absorption layer). Therefore, according to some embodiments, the duration of the pump pulse can be selected to not exceed or substantially not exceed the thermal expansion time of the absorption layer. According to some such embodiments, where the absorption layer is silicon-based, the duration of the pump pulse can be less than about 5 psec (picoseconds), about 3 psec, or even about 1 psec. Each possibility corresponds to a different embodiment.

[0247] It is further noted that the position of the absorption layer can be controllable. For example, the sample may include a first portion of a first material and a second portion of a second material, wherein the second portion is located within the first portion. The absorption layer can then be controllably positioned in the second portion by selecting a pump pulse characterized by wavelength, such that the absorption length of the pump pulse in the first material is much greater than its absorption length in the second material.

[0248] Depending on the composite structure of the sample, the absorber layer may or may not be included in the target region. Different possible locations of the absorber layer within the sample are shown in Figure 1. Figure 4 , Figure 5 , Figures 6D to 6G ,as well as Figures 7C to 7E The diagram shows that, according to some embodiments, the absorption layer may be formed as, or included in, a dissimilar component embedded in or on the target region. The embedded component may be characterized by absorption behavior different from that of the remainder of the target region (e.g., different absorption lengths and / or different correlations of absorption lengths with polarization), thereby allowing selective heating of the embedded component.

[0249] According to some embodiments, the absorption layer is silicon-based, and the wavelength of the pump pulse is in the ultraviolet range (i.e., below 360 nm). According to some embodiments, the absorption layer is metallic, and the wavelength of the pump pulse can also be selected from the visible range.

[0250] The pump pulse heats the absorption layer within it, causing it to expand and resulting in one or two acoustic pulses in the sample. More specifically, when the absorption layer is the outermost layer of the sample (on which the pump pulse is projected directly), a single acoustic pulse is generated that propagates away from the absorption layer and into the sample in a direction perpendicular to the absorption layer. When the absorption layer is the innermost layer of the sample, two acoustic pulses are generated that propagate away from the absorption layer, perpendicular to the absorption layer, and in opposite directions.

[0251] The lateral range of the acoustic pulse can be selected to be greater than the lateral range of the lateral structural features. More precisely, to fully probe the target region, the lateral size of the acoustic pulse can be selected to be greater than or approximately equal to the lateral size of the target region. Furthermore, the lateral size of the acoustic pulse is determined by the beam diameter of the pump pulse. As explained above, this sets a lower limit on the wavelength of the pump pulse, because optical diffraction limitations restrict the diameter of the laser beam to be greater than approximately λ / 2, where λ is the wavelength of the laser beam. Subsequently, in sub-operation 320a, the wavelength of the pump pulse is selected to be greater than or at least approximately equal to twice the lateral range of the lateral structural features.

[0252] According to some embodiments, the depth dimension (also known as the "longitudinal dimension") of the target area is determined by the propagation direction of the acoustic pulse within the target area.

[0253] In sub-operation 320b, the wavelength of the probe pulse can be selected to allow the probe pulse to traverse the target region, thereby allowing complete detection of the target region. More specifically, the wavelength of the probe pulse can be selected such that the absorption length of the probe pulse within the target region is greater than (or at least approximately equal to) the range of the target region along the propagation direction of the probe pulse within the target region.

[0254] According to some embodiments, the wavelength of the probe pulse is selected to be at least about twice the lateral range of the lateral structural feature.

[0255] According to some embodiments, different probe pulses configured to scatter the acoustic pulse at different depths can be characterized by different wavelengths, and more generally by waveform and / or polarization. When the target region comprises different types of layers (e.g., lateral layers) characterized by different refractive indices and / or sound velocities (e.g., due to different material compositions or internal geometries), this property of probe pulse wavelength (or waveform and / or polarization) to scattering depth can be achieved using a probe modulator (e.g., probe modulator 232). In particular, this allows for selective proberization of each type of layer.

[0256] As used herein, according to some embodiments, the absorption length of a sample (e.g., a material body or a composite structure comprising multiple parts made of different materials and optionally characterized by different geometries) is defined as the distance over which the intensity of a light beam entering the sample decreases to approximately 1 / e (≈63%) of the intensity of the beam when it enters the body. As used herein, according to some embodiments, the terms "sample" and "sample" may be used interchangeably.

[0257] It should be noted that the absorption length of a composite sample depends not only on the absorption length of each of the multiple parts that make up the sample, but also on the geometry of the multiple parts and their spatial arrangement relative to each other. Thus, for example, a sample comprising alternating layers of two different materials, such that a sample in which each of the two materials is transparent or substantially transparent to radiation in a continuous wavelength range, may reflect radiation at a specific wavelength within that range in any way.

[0258] According to some embodiments, a linearly polarized probe pulse can be generated. Specifically, according to some embodiments, and as described below... Figure 6G As described, linearly polarized probe pulses can be used to increase measurement sensitivity along a selected lateral direction. According to some embodiments, the polarization of the pump pulse can be selected to increase its absorption in the absorption layer. In this regard, the geometry of the absorption layer can play an important role. For example, when the absorption layer comprises multiple parallel stripes (e.g., as in...), Figures 7C to 7E (As depicted in the text), the polarization of the pump pulse can be selected to be parallel to the strip.

[0259] According to some embodiments, particularly in embodiments where the sample comprises a composite lateral structural feature consisting of multiple identical lateral structural features (whether periodic or not), the lateral structural features constituting these multiple identical lateral structural features can be simultaneously and deeply profiled. In such embodiments, the target region is selected to include multiple lateral structural features by correspondingly selecting the beam width of the pump pulse (which defines the lateral area of ​​the target region). The beam width of the probe pulse can be set to the beam width of the pump pulse, thereby ensuring complete detection of the lateral range of the target region (such that all lateral structural features among the multiple identical lateral structural features are detected). Considering optical diffraction limitations, the wavelength of the pump pulse is selected to be at least about twice the lateral range of the composite lateral structural feature. Similarly, the wavelength of the probe pulse is selected to be at least about twice the lateral range of the composite lateral structural feature. The obtained multiple measured signals then exhibit an average depth correlation of parameters characterizing the lateral structural features among the multiple lateral structural features. According to some such embodiments, the composite lateral structural feature is periodic.

[0260] According to some embodiments, in operation 320, the temperature of the sample can be adjusted to ensure that at the beginning of each of the m implementations, the temperature of the sample is the same, and optionally equal to a predetermined temperature.

[0261] In operation 330, multiple measured signals {M} are obtained in the m implementations of operation 320. r ′(s)} r It can be demodulated and combined to obtain a single (combined) signal ES′(s). Here, s represents the depth within the target region. M r Let ′(s) represent the measured signal obtained for the probe pulse that scatters away from the acoustic pulse at depth s. Note that for a given depth s, {M r ′(s)} r It can typically include multiple measured signals obtained at depth s.

[0262] As a non-restrictive and purposefully simplified example intended to make the discussion clearer, consider multiple measured signals {M}. r "(s)} r＝1,2 This includes two measured signals M1″(s) A ) and M2″(s B (Typically, multiple measured signals can include any number between approximately 10 and approximately 1000 measured signals.) Assume the measured signal is M1″(s... A ) and M2″(s B ) have been respectively passed through depth s A and s B >s A The signal is obtained by scattering the probe pulse away from the acoustic pulse. This is achieved from two measured signals M1″(s). A ) and M2″(s B Two extraction signals can be obtained. and (For example, using a lock-in amplifier), where background signals and noise are suppressed, as described above in the description of system 200. The two extracted signals can be combined into a single combined signal ES″(s), where for s... A -1 / 2·Δs≤s≤s A +1 / 2·Δs, And for s B -1 / 2·Δs≤s≤s B +1 / 2·Δs, Here, Δs can correspond to the layer thickness detected by the probe pulse at each of the two depths. Additionally, it is implicitly assumed that s A +1 / 2·Δs≤s B -1 / 2·Δs.

[0263] The depth s at which scattering occurs can be related to the scattering time based on the time delay Δt (of the probe pulse relative to the pump pulse) and the acoustic pulse t. F The formation time is related to the propagation speed of the sound pulse in the target area. Formation time t F This is the time it takes for an acoustic pulse to form once the absorption layer has been irradiated by the pump pulse. The propagation speed of the acoustic pulse is equal to the speed of sound, v. 声 (As explained in detail below, in a non-uniform medium, the speed of sound can depend on depth, and in this case, its functional dependence on depth can be considered.)

[0264] According to some embodiments, s itself may depend linearly or substantially linearly on the time delay Δt. According to some such embodiments, for example, the absorption layer is located adjacent to the target region but deeper within the sample compared to the target region (e.g., as in...). Figures 6D to 6G (as shown in the diagram), s = Dv 声 ·(Δt-t F D is the depth dimension or longitudinal extent of the target region. Therefore, for the minimum time delay (i.e., Δt = t), F ), s = D, and the target region is detected at the maximum depth. For the maximum delay time (i.e., Δt = t), F +D / v 声 ), s=0 and the target region is detected at zero depth. According to some alternative embodiments, for example, the absorption layer forms the shallowest layer of the target region (e.g., as shown in the figure). Figures 6D to 6G (as shown in the diagram), s = v 声 ·(Δt-t F Therefore, for the minimum delay time (i.e., Δt = t), F ), s = 0 and the target region is detected at the minimum depth. For the maximum delay time (i.e., Δt = t), F +D / v 声 ), s = D and the target area was detected at the maximum depth.

[0265] According to some embodiments, as part of the analysis in operation 330, the combined signal can be compared with another signal measured in the absence of an acoustic pulse (i.e., when no pump pulse is projected onto the sample). The comparison allows for the isolation of the acoustic pulse's contribution to the combined signal, and thereby facilitates the extraction of Brillouin oscillations resulting from the interaction between the probe pulse and the acoustic pulse.

[0266] According to some embodiments, the generation of an acoustic pulse in the target region (attributed to the expansion of the absorbing layer) may be accompanied by a change in the refractive index of the target region (or a portion thereof) attributable to a thermo-optical effect, i.e., a change in the refractive index of the medium attributable to a change in temperature within the medium. The change in reflectivity caused by the acoustic pulse and attributable to the thermo-optical effect, and particularly its relative intensity, depend on the physical properties of the medium.

[0267] According to some embodiments, operation 330 may include a sub-operation in which the thermo-optical contribution to the combined signal can be removed or substantially removed. Specifically, according to some embodiments, the thermo-optical contribution to the combined signal itself manifests as a slowly varying contribution to the addition of Brillouin oscillations (attributed to elasto-optical effects). That is, the Brillouin frequency is much higher than the frequency associated with the contribution of the thermo-optical effect to the combined signal. Therefore, the thermo-optical contribution to the combined signal can be identified and removed, for example, by smoothing the combined signal (i.e., by averaging short segments of the signal such that each segment includes a small number of (Brillouin) oscillations). This can be particularly useful when the target region is a silicon-based semiconductor, since in silicon-based semiconductors, the thermo-optical effect can be much stronger than the elasto-optical effect.

[0268] According to some embodiments, computer simulation can be used to model Brillouin oscillations, or even a single Brillouin oscillation, which would be observed if the ideal (i.e., perfectly manufactured) sample realization method 300 were performed. (Note that this is the case when the longitudinal extent of the target region is comparable to the Brillouin wavelength, for example, in the case of a fin field-effect transistor where only a single Brillouin oscillation can be observed in depth profiling.) The degrees of freedom in selecting the physical parameters of the characterization settings allow for the “manual” elimination of the contribution of thermo-optical effects, thus avoiding signal processing operations that would otherwise distinguish Brillouin oscillations from thermo-optical contributions. Additionally, according to some embodiments, computer simulation can also be used to model various types of defects in samples and systems and their associated Brillouin oscillations.

[0269] Furthermore, a scanning electron microscope (SEM) can be used to scan the deeply profiled sample to obtain the actual (or true) structure of the target region. More specifically, the target region of a deeply profiled sample (using the methods of this disclosure) can be cut into sufficiently thin layers, and each layer can be scanned by (SEM) to obtain its actual structure. Thus, the Brillouin oscillations (and more generally, multiple measured signals) obtained from different samples can be correlated with the actual structure of their respective target regions.

[0270] Using machine learning tools, a measurement data analysis module (such as measurement data analysis module 208) can be taught to extract deep correlations of one or more parameters characterizing the detected lateral structural features from observed Brillouin oscillations. Supervised teaching can be employed using observed and / or simulated Brillouin oscillations, which have corresponding structures as measured or simulated lateral structural features, respectively (e.g., using SEM). Such Brillouin oscillation-lateral structural feature pairs, or pairs of similar types belonging to similar settings, may also be available from existing databases (e.g., online databases) of measured and / or simulated signals in similar settings (i.e., similar samples and systems).

[0271] In a homogeneous medium, the Brillouin frequency f B By f B =(2·v 声 ·n) / λ 探测 Given. Here, n is the refractive index and λ 探测 It is the wavelength of the probe pulse. In inhomogeneous media, according to some embodiments, the Brillouin frequency can be determined by both the material composition and geometry of the structure. That is, according to some embodiments, the speed of sound v 声 The refractive index n can generally be determined by the "effective speed of sound" v at depth s. eff (s) and “effective refractive index” n eff (s) substitution. Brillouin frequency, f B (s)=(2·v eff (s)·n eff (s)) / λ 探测 Therefore, it generally depends on the depth s. The extracted signal can thus be obtained by taking O... B (s)=A(s)·sin(2π·f B (s)·s+φ1) form. Therefore, by extracting the signal O B (s) obtain f B (s) can be estimated v eff (s) and n eff (s). v eff (s) and n eff (s) can then be associated with one or more parameters characterizing the lateral structural features being attempted to be deeply dissected. According to some embodiments, f B (s) can be directly related to lateral structural features (e.g., directly related to the average diameter of holes in an array of vertical holes, as described below). Figures 9A to 9E (As described in the description).

[0272] According to some embodiments, regression analysis can be used to extract deep correlations of one or more lateral structural features from multiple measured signals.

[0273] According to some embodiments, multiple operations 320 can be performed with respect to different preparations of the pump pulse and / or probe pulse. The different preparations may differ from each other in one or more of the following: the wavelength, power, waveform, and / or polarization of the pump pulse, and / or the wavelength, power, waveform, and / or polarization of the probe pulse. After preprocessing (e.g., demodulation using a lock-in amplifier, smoothing the thermo-optical contribution to the measured signal), multiple measured signals according to each preparation can be jointly analyzed to determine the depth correlation of at least one parameter characterizing the lateral structural features.

[0274] According to some embodiments, where the sample is a wafer, operation 320 can be repeated with respect to different locations on the wafer. Multiple measurement signals can be analyzed to obtain information about processing variations across the wafer.

[0275] According to some of these embodiments, the same location on different grains (the variation between grains) can be detected and the multiple measured signals obtained can then be analyzed to obtain a large-scale map of the processing variation across the wafer.

[0276] According to some embodiments, different locations on the same grain known to be characterized by the same lateral structural features up to manufacturing defects can be detected, and multiple measured signals obtained can be subsequently analyzed to obtain a graph of processing variations across grains (variations within the grain) or one or more regions thereof, as described above. Figure 2 As described in the description.

[0277] Figure 4 and Figure 5 Each of these schematically depicts, based on some embodiments, additional possible configurations (spatial arrangements) of the target region and absorption layer within the sample. See also Figure 4 According to some embodiments, a sample 400 (e.g., a semiconductor component) undergoing in-depth profiling is schematically depicted. This differs from that included in target region 110. Figures 1A to 1C The absorber layer 120 of sample 100 and the absorber layer 420 of sample 400 are not included in the target region 410 of sample 400.

[0278] although Figure 4 To facilitate the description, a two-dimensional view of the target region 410, cut along the zx plane up to the manufacturing defect, is provided, but it is assumed that the target region 410 exhibits rotational symmetry along an axis parallel to the z-axis (making the target region 410 cylindrical). Therefore, Figure 4 The target area 410 was effectively and completely depicted.

[0279] More specifically, sample 400 includes a target region 410, an outer region 430, and an inner region 440. The target region 410 is located between the outer region 430 and the inner region 440. The absorber layer 420 is located within the inner region 440 adjacent to the target region 410. Dashed line B4 indicates the circumferential boundary of the target region 410.

[0280] To make the discussion more specific and thereby facilitate the description, the target region 410 is depicted as comprising two sub-regions: an outer target sub-region 410a and an inner target sub-region 410b surrounded by the outer target sub-region 410a. The inner target sub-region 410b may correspond to a first medium characterized by a first refractive index, while the outer target sub-region 410a may correspond to a second medium characterized by a second refractive index different from the first refractive index.

[0281] exist Figure 4 In the diagram, we observe that the width w4(z) of the target's internal sub-region 410b increases abruptly (i.e., discontinuously) along the negative z-axis. The rotational symmetry of the target region 410 implies that the target's internal sub-region 410b is a circular stepped pyramid. Therefore, w4(z) corresponds to the (depth-dependent) diameter of the target's internal sub-region 410b. The lateral structural features consist of a change from the first medium to the second medium along any radial direction, which, when starting from axis A4, is perpendicular to the rotational symmetry axis A4 of the target region 410 (axis A4 is parallel to the z-axis). As defined above, the lateral extent of the lateral structural features can be determined by max... z [w4(z)] is given.

[0282] A pump pulse 423 is also depicted projected onto the outer surface 424 of the outer region 430 (and sample 400). The pump pulse 423 is configured to penetrate into the sample 400 and reach the absorption layer 420 by propagating through the outer region 430 and the target region 410. In particular, the outer region 430 and the target region 410 may be transparent or substantially transparent to the pump pulse 423. The pump pulse 423 is further configured to be absorbed in the absorption layer 420 and thereby heat and expand the absorption layer 420. The expansion of the absorption layer 420 generates an acoustic pulse 425 that propagates into the target region 410 in the direction of the negative z-axis. A second acoustic pulse (not shown) may propagate in the direction of the positive z-axis within the inner region 440.

[0283] It also describes the probe pulse 427 projected onto the sample 400 with a controllable time delay from the pump pulse 423. (Therefore, Figure 4 (This should be understood as illustrative and not representing a single moment.) The probe pulse 427 is configured to penetrate into and propagate within the sample 400 so as to scatter away from the acoustic pulse 425 at a controlled depth within the target region 410.

[0284] The scattered component 431 of the probe pulse 427, which is backscattered (Brillouin) away from the acoustic pulse 425, is further indicated. The scattered component 431 can be detected by the detector 432 to generate the corresponding measured signal.

[0285] See Figure 5 According to some embodiments, a sample 500 undergoing depth analysis is schematically depicted. This differs from the outer surface 124 of the adjacent sample 100, which is positioned... Figures 1A to 1C The target region 110 of sample 100 and the target region 510 of sample 500 are completely embedded in sample 500 (and therefore, not located adjacent to the outer surface 524 of sample 500).

[0286] although Figure 5 To facilitate the description, a two-dimensional view of the target region 510, intercepted along the zx plane, is provided, but it is assumed that the target region 510 is uniform along the y-axis until the manufacturing defect is uniform. Therefore, Figure 5 The target area 510 was effectively and completely depicted.

[0287] More specifically, sample 500 includes a target region 510, an outer region 530, and an inner region 540. The target region 510 is located between the outer region 530 and the inner region 540. The absorption layer 520 is located within the target region 510 adjacent to the outer region 530. Dashed line B5 indicates the first boundary of the target region 510, and dashed line B5' indicates the second boundary of the target region 510. To make the discussion more specific and thereby facilitate the description, the target region 510 is depicted as comprising three adjacent sub-regions: a first lateral sub-region 510a, a second lateral sub-region 510b, and an intermediate sub-region 510c located between the first lateral sub-region 510a and the second lateral sub-region 510b. The width w5(z) of the intermediate sub-region 510c increases in the direction of the positive z-axis.

[0288] A pump pulse 523 is also depicted projected onto the outer surface 524 of the outer region 530 (and sample 500). The pump pulse 523 is configured to penetrate into the sample 500 and reach the absorption layer 520 by propagating through the outer region 530. In particular, the outer region 530 may be transparent or substantially transparent to the pump pulse 523. The pump pulse 523 is further configured to be absorbed in the absorption layer 520 and thereby heat and expand the absorption layer 520. The expansion of the absorption layer 520 generates an acoustic pulse 525 that propagates in the direction of the positive z-axis into the target region 510. A second acoustic pulse (not shown) may propagate in the direction of the negative z-axis within the outer region 530.

[0289] A probe pulse 527, which is projected onto sample 500 with a controllable time delay, is also described. (Therefore, Figure 5(This should be understood as illustrative and not representing a single moment.) The probe pulse 527 is configured to penetrate into and propagate within the sample 500 so as to scatter away from the acoustic pulse 525 at a controlled depth within the target region 510. The probe pulse 527 may be further configured to undergo relatively small scattering by a second acoustic pulse propagating within the outer region 530 (i.e., the total cross-section of the second acoustic pulse scattered away may be significantly smaller than the total cross-section of the acoustic pulse 525 scattered away). For example, the waveform of the probe pulse 527 may be selected such that the probe pulse 527 is focused within the target region 510 but defocused within the outer region 520.

[0290] The scattered component 531 of the probe pulse 527, which is backscattered (Brillouin) away from the acoustic pulse 525, is further indicated. The scattered component 531 can be detected by the detector 532 to generate a corresponding measurement signal.

[0291] Figures 6A to 6F A sample 600 undergoing in-depth analysis is schematically depicted according to some embodiments. See also Figure 6A Sample 600 is depicted with its front portion removed to better reveal its internal structure. Sample 600 includes a structure 602 positioned on a body 604. Structure 602 includes a (air) aperture 608 projected therein. According to some embodiments, aperture 608 may be projected into structure 602 from its top (as depicted in the figures) outer surface 610. Structure 602 may be characterized by a first (effective) refractive index and body 604 may be characterized by a second refractive index different from the first refractive index.

[0292] Due to the presence of the opening 608, structure 602 includes a plurality of transverse structural features constituting a composite transverse structural feature. According to some embodiments, and as in... Figure 6A The image depicts a composite lateral structural feature forming a repeating pattern. Specifically, the holes 608 are arranged in a periodic two-dimensional array. According to some embodiments of this, the two-dimensional array is rectangular, with the holes 608 arranged in rows and columns parallel to the x-axis and y-axis, respectively.

[0293] More specifically, the lateral structural characteristics are associated with each of the apertures 608, and these characteristics are constituted by the change from air to solid that occurs along any radial direction perpendicular to the longitudinal axis when starting from the longitudinal axis of the aperture. (The longitudinal axis extends parallel to the z-axis.) The two longitudinal axes A of apertures 608a and 608b are... a and vertical axis A b In respectively Figure 6B The instructions are in accordance with the central government.

[0294] For the sake of clarity, in the following description, it is assumed that each of the holes 608 is projected longitudinally into the structure 602 and is characterized by an elliptical transverse cross-section whose area decreases with depth. That is, each of the holes 608 can be represented by the (conjugate) diameter d of the width of the hole 608 along the x-axis and y-axis, respectively. x (s) and d y (s) Characterization. Two such diameters d of holes 608a and 608b. x ′(s) and d x "(s) in Figure 6A The location s = 0 is indicated, i.e., on the top outer surface 610. Here, s is the depth within the sample 600. (Generally, s = z + k, where k is a constant. If the coordinate system is chosen such that the xy plane coincides with the top outer surface 610, then k = 0 and s = z.) Figure 6B The diagram also indicates the diameters d of holes 608a and 608b at depths s = s′ and s = s″, respectively. x (a) (s) and d x (b) (s).

[0295] For the "lowest order," the depth dependence of the transverse structural features can be parameterized by the depth dependence of the transverse cross-sectional area (of the hole). For greater accuracy, the depth dependence of two parameters can be estimated, namely the two conjugate diameters of the ellipse, as detailed below. If even greater accuracy is required, in principle, one can also attempt to obtain the depth dependence of additional parameters that parameterize each deformation, which may be depth-dependent. For example, parameters characterizing the inclination of the hole's axis of symmetry (which may depend on depth), the deviation of the spacing between adjacent holes (from the design-specified spacing), and so on.

[0296] According to some embodiments, in order to obtain the depth correlation of multiple parameters characterizing the transverse structural features, operation 320 can be implemented with respect to different preparations of the pump pulse and / or probe pulse. For example, one or more operations 320 can be performed using probe pulses polarized parallel to the x-axis, and one or more operations 320 can be performed using probe pulses polarized parallel to the y-axis. In this way, the average depth correlation of each of the two conjugate diameters characterizing the elliptical cross-section of the hole 608 can be obtained.

[0297] Figure 6B A partial cross-sectional view of sample 600 is provided according to some embodiments. The cross-section cuts sample 600 along a plane parallel to the zx plane. According to some embodiments, and as in Figures 6B to 6F As depicted, structure 602 can be a layered structure comprising multiple layers 612 stacked (i.e., positioned) on top of each other. (Layers 612 are not in...) Figure 6A(See illustration in the middle.) According to some embodiments, layer 612 may include two types of layers: layer 612a and layer 612b, which are alternately positioned on top of each other. Layers 612a and 612b may be made of different materials. According to some embodiments, sample 600 may be a V-NAND (i.e., vertical NAND) stack, wherein structure 602 is mounted on a silicon substrate formed by body 604. As a non-limiting example, according to some such embodiments, layer 612a (including the outermost layer) may be made of silicon oxide (SiO2) and layer 612b may be made of silicon nitride (e.g., Si3N4).

[0298] Figures 6C to 6F The diagrams schematically depict the target region 624 (in) according to method 300. Figure 6A The diagram shows four consecutive stages in the depth profile (illustrated in the middle and depicted by double dashed lines). See also Figure 6C According to some embodiments, pump pulse 623 is illustrated as being projected onto the top outer surface 610. Pump pulse 623 is configured to penetrate into and propagate within structure 602 to reach body 604. Pump pulse 623 is further configured to be absorbed by body 604. Pump pulse 623 also... Figure 6A The instructions are in accordance with the central government.

[0299] See Figure 6D The pump pulse 623 is absorbed in the absorption layer 618 (which forms part of the body 604) located adjacent to the structure 602. The thickness of the absorption layer 618 is determined by the absorption length of the pump pulse 623 within the body 604. (As shown by...) Figure 6B The double-headed arrow e6 indicates that heating the absorber layer 618 causes it to expand.

[0300] See Figure 6E The expansion of the absorption layer 618 results in a (first) acoustic pulse 625a that propagates within the structure 602 in the direction of the negative z-axis. A second acoustic pulse 625b can propagate within the body 604 in the direction of the positive z-axis.

[0301] See Figure 6F According to some embodiments, a probe pulse 627 is illustrated as being projected onto a top outer surface 610. The probe pulse 627 is configured to penetrate into the structure 602 and propagate therein in the direction of the positive z-axis. That is, the probe pulse 627 is configured such that the structure 602 is transparent or at least semi-transparent to the probe pulse 627, at least when there is no interference. An acoustic pulse 625a is locally present in a sub-region within the structure 602 such that the sub-region is opaque to the probe pulse 627. More precisely, the probe pulse 627 is further configured to undergo Brillouin scattering away from the acoustic pulse 625a. The (backward) scattering component 631 of the probe pulse 627 propagates away from the acoustic pulse 625a in the direction of the negative z-axis.

[0302] See you again Figure 6A The target region 624 is included in the structure 602. The target region 624 is cylindrical in shape, and its diameter is defined by the beam diameter of the pump pulse 623. According to some embodiments, in order to fully probe the target region 624, the beam diameter of the probe pulse 627 may be selected to be equal to or substantially equal to the diameter of the pump pulse 623. The target region 624 thus constitutes a part of the structure 602 that undergoes in-depth profiling.

[0303] Note that the target region 624 includes multiple holes from the aperture 608, and specifically, multiple lateral structural features. Since the holes in the target region 624 are together depth-profiled, multiple measurement signals obtained when the target region 624 is subjected to depth profiling according to method 300 collectively characterize the depth correlation of the multiple lateral structural features (included in the target region 624). That is, the average depth correlation of the multiple measured signals characterizes a parameter that characterizes the lateral structural feature. Specifically, from the multiple measured signals, the lateral cross-sectional area of ​​the aperture or the average depth correlation of two conjugate diameters characterizing the lateral cross-section of the aperture can be obtained. Alternatively, the beam diameter of the probe pulse 627 (assuming it is equal to the beam diameter of the pump pulse 623) defines the target region (i.e., the target region 624) including multiple holes and thus multiple lateral structural features (which together define a composite lateral structural feature). The lateral extent of the composite lateral structural feature is equal to the beam diameter of the probe pulse 627.

[0304] According to some embodiments, where the apertures 608 are arranged in a two-dimensional array, the probe pulses 627 can be linearly polarized along a direction perpendicular to the z-axis. More specifically, according to some embodiments, where the apertures 608 are arranged in a rectangular array as described above to increase measurement sensitivity along the y-axis, the probe pulses 627 can be polarized parallel to the x-axis. Similarly, to increase measurement sensitivity along the x-axis, the probe pulses 627 can be polarized parallel to the y-axis. As illustrated in the simulation results section below, polarizing the probe pulses parallel to the x-axis results in a non-uniform intensity distribution of the probe pulses within the target region, where the intensity is greatest along the columns of the apertures. In contrast, polarizing the probe pulses parallel to the y-axis results in a non-uniform intensity distribution of the probe pulses within the target region, where the intensity is greatest along the rows of the apertures.

[0305] Figure 6G A depth profile of the target region 624 is depicted according to some embodiments. The depicted depth profile is configured in... Figures 6A to 6F The specific embodiment of the depth profile depicted is shown. The probe pulse 627' is depicted, which is a specific embodiment of the probe pulse 627. As shown by the polarization arrow P... yThe indication is that the probe pulse 627' is polarized along the y-axis. This allows for increased measurement sensitivity along the x-axis, making it possible to obtain the average depth correlation variation of the aperture's x-width (i.e., the average function d over the aperture). x (s) to a greater degree of accuracy. (Therefore, for example, if only holes 608a and 608b are probed, the obtained depth correlation constitutes an average over the x-width of holes 608a and 608b).

[0306] Still Figure 6G The text indicates a (first) acoustic pulse 625a', a second acoustic pulse 625b', and a scattering component 631', which are specific embodiments of acoustic pulse 625a, second acoustic pulse 625b, and scattering component 631. The scattering component 631' is also polarized along the y-axis.

[0307] Despite Figures 6A to 6G In the present invention, the diameter of hole 608 is described as decreasing linearly with depth. Those skilled in the art will understand that the methods and systems disclosed herein can be applied to detect other hole geometries, such as, for example, circular hole geometries, where the change in hole diameter (with increasing depth) is non-monotonic (e.g., increasing first and then decreasing).

[0308] Figures 7A to 7E A sample 700 undergoing in-depth profiling is schematically depicted according to some embodiments. The sample 700 includes a base portion 704 and a plurality of fins 708 positioned on the base portion 704. Each of the fins 708 forms an elongated ridge-like structure projecting from the base portion 704. According to some embodiments, and as in... Figures 7A to 7E As depicted, fins 708 have the same shape and are arranged parallel to each other.

[0309] Sample 700 includes multiple transverse structural features forming a composite transverse structural feature. According to some embodiments and as shown in... Figures 7A to 7E As depicted, the composite lateral structural features form a repeating pattern (i.e., periodically repeating in the x-axis direction). For each of the fins 708, as it traverses the fin in a lateral direction perpendicular to the elongation dimension (i.e., parallel to the x-axis), the associated lateral structural feature is constituted by a change from air to fin and back to air.

[0310] To facilitate description and make the argument more specific, in Figures 7A to 7EIn this text, the width w7(z) of fin 708 is depicted as decreasing with distance from the base portion 704. (However, those skilled in the art will understand that other geometries are possible.) w7(z) constitutes a parameter characterizing the depth relevance of the lateral structural features in the sense defined above herein. If greater accuracy is required, in principle, an attempt can be made to obtain additional parameters of depth relevance, such as parameters characterizing deviations from design specifications (attributed to manufacturing defects) of the slopes of the right and left walls of the fin. The lateral extent C7 of the fin may correspond to the maximum width of the fin, which is within... Figures 7A to 7E The width of the fin corresponds to the width of the fin at the base of its adjacent base portion 704. That is, C7 = max z [w7(z)].

[0311] According to some embodiments, and as in Figures 7A to 7E As depicted, multiple fins 708 constitute a target region 710 that is deeply profiled. According to some of these embodiments, the sample 700 may be a fin field-effect transistor (FinFET). In this embodiment, the sample 700 may be made of silicon, germanium silicon, or other suitable semiconductor materials.

[0312] Figures 7B to 7E Cross-sectional views of sample 700, depicting four consecutive stages of depth profiling according to method 300, are provided. The cross-sections are cut along a plane parallel to the zx plane. See also... Figure 7B According to some embodiments, the pump pulse 723 is illustrated as being projected onto the sample 700 (with the fin 708 positioned on the side of the sample 700 thereon).

[0313] See also Figure 7C The pump pulse 723 is configured to be absorbed in fin 708. More specifically, the pump pulse 723 is configured to be absorbed in (lateral) absorption layers 712. Each of the absorption layers 712 constitutes a top stripe of the corresponding fin from fin 708. For example, absorption layer 712a from absorption layer 712 constitutes a top stripe of fin 708a from fin 708. The thickness of the absorption layer 712 may be determined (or primarily determined) by the absorption length of the pump pulse 723 within the fin. The absorption length, in turn, depends at least on the wavelength (and polarization angle) of the pump pulse 723. Figure 7C As indicated by the double-headed arrow e7, heating the absorber layer 712 causes it to expand.

[0314] See Figure 7D The expansion of the absorption layer 712 results in the formation of acoustic pulses 725. Each of the acoustic pulses 725 propagates away from the absorption layer toward the base portion 704 within the corresponding fin (from fin 708). For example, acoustic pulse 725a (from acoustic pulse 725) propagates within fin 708a in the direction of the negative z-axis.

[0315] See Figure 7E According to some embodiments, a probe pulse 727 is illustrated as being projected onto a sample 700 (on the side of the sample 700 from which the fin 708 projects). The probe pulse 727 is configured to penetrate into the fin 708 and propagate therein in the direction of the positive z-axis. That is, the probe pulse 727 is configured such that the fin 708 is transparent or at least translucent to the probe pulse 727, at least when undisturbed. Acoustic pulses 725 are locally present in corresponding sub-regions within the fin 708, making these sub-regions opaque to the probe pulse 727. More precisely, the probe pulse 727 is further configured to undergo Brillouin scattering away from the acoustic pulse 725. The (backward) scattering component 731 of the probe pulse 727 propagates away from the acoustic pulse 725 in the direction of the negative z-axis.

[0316] Note that, for the configuration of pump pulse 723 and probe pulse 727 described above, fin 708 is also probed simultaneously. Therefore, when the target region 710 undergoes depth profiling according to method 300, the multiple measurement signals obtained using the configuration of pump pulse 723 and probe pulse 727 described above collectively characterize the depth correlation of the lateral structural features included in the target region 710 (in... Figures 6A to 6G (In the sense described above). Referring to sample 700, at least the average depth correlation of the width of fin 708 can be extracted from these multiple measured signals.

[0317] According to some embodiments, the pump pulse 723 and / or the probe pulse 727 can be linearly polarized along the elongation dimension of the fin 708 (i.e., along the y-axis), thereby increasing measurement efficiency. This selection of the polarization of the pump pulse 723 increases its absorption in the absorption layer 710 and also in the sidewall 716 of the fin 708 (also... Figure 7A The absorption in the fin (numbered in the middle) is minimized. Furthermore, as illustrated below in the simulation results section, the above choice of polarization for the probe pulse 727 maximizes its penetration into the fin 708. In contrast, if the probe pulse 727 is polarized perpendicular to the elongated dimension of the fin 708 (i.e., along the x-axis), the fin 708 is essentially transparent to it. That is, in the latter case, essentially all radiation is concentrated in the space between the fins outside the fin 708.

[0318] The polarization direction of pump pulse 723 is in Figure 7A and Figure 7B The polarization direction of the probe pulse 727 is indicated by the polarization arrow Qy. Figure 7E The polarization direction is indicated by the polarization arrow Qy′, which is also the polarization direction of the scattering component 731.

[0319] Simulation results

[0320] Figure 8The extracted signal, obtained through computer simulation of an implementation of a method 300 for V-NAND stacking, is depicted according to some embodiments. The V-NAND stack is constructed using a method similar to sample 600 (…). Figure 6A The aperture profile of aperture 608 (described in Figure 6H) is characterized. More specifically, the (normalized) extracted signal E obtained when a series of probe pulses are scattered away from the acoustic pulse at different depths within the sample is depicted. E = ΔI / I, where I represents the intensity of the measured signal and ΔI represents the deviation from the baseline of the measured signal (caused by Brillouin scattering). As described above... Figure 6A As described in Figure 6H, the acoustic pulse is generated by a uniformly prepared pump pulse. The horizontal axis measures the time t from the generation of the acoustic pulse, such that the larger t is, the smaller the scattering depth. The maximum scattering depth corresponds to t = 0 (at which point the acoustic pulse penetrates, for example, from the absorption layer 618 into the structure 602). The end of the time scale indicates the time it takes for the acoustic pulse to reach the outer surface of the sample to which the probe pulse is incident (e.g., the top outer surface 610). Multiple Brillouin oscillations are observed, where the amplitude of the oscillations decreases with t, i.e., increases with depth.

[0321] Using frequency estimation techniques, such as short-time Fourier transform or sinusoidal fitting over short time intervals, the time correlation of Brillouin frequencies can be obtained from the extracted signal. This, in turn, allows for the acquisition of local Brillouin frequencies, i.e., Brillouin frequencies that vary with depth *s*. Note that when the speed of sound is constant throughout the target region, the relationship between time *t* and depth *s* is linear.

[0322] Figures 9A to 9E According to some embodiments, computer simulation results of depth profiling of five samples using method 300 are provided respectively. The five samples are respectively... Figures 10A to 10E The five samples are depicted in the diagram. Each of them comprises a uniform array of vertically extending holes. The samples differ from one another in terms of the hole profiles (i.e., shapes). Each of the five modeled samples is similar to sample 600 (up to the different hole profiles) and therefore can correspond to a possible design of V-NAND stacking, or to specific possible distortions in such a design due to manufacturing defects. For example, Figure 10A The hole profile can represent the possible specific design of V-NAND stacking, while Figure 10E This can indicate its specific potential distortion. Or, for example, Figure 10B The hole profile can represent the possible specific design of V-NAND stacking, while Figure 10D This can represent its potential specific distortion. The vertical axis parameterizes, for example, the depth *s* measured from the top of the hole. The horizontal axis *x* parameterizes the width of the hole. Figures 9A to 9E In each of them, only half of the hole is shown in the diagram, which implicitly suggests that the hole exhibits rotational symmetry about a vertical axis (as in...). Figures 10A to 10E(Depicted in Chinese).

[0323] See Figure 10A The figure shows a cross-sectional view of sample 1000a. The diameter of the pore 1008a within sample 1000a is constant (i.e., it does not change with depth). See also Figure 10B The figure shows a cross-sectional view of sample 1000b. The diameter of the pore 1008b within sample 1000b decreases at a constant rate with depth. See also... Figure 10C The figure shows a cross-sectional view of sample 1000c. The diameter of the pore 1008c within sample 1000c decreases with depth at a first rate for depths less than a threshold depth (not indicated), and decreases with depth at a second rate for depths greater than the first depth. The first rate is greater than the second rate. See also... Figure 10D The figure shows a cross-sectional view of sample 1000d. The diameter of the pore 1008d within sample 1000d decreases with depth at a first rate for depths less than a threshold depth (not indicated), and at a second rate for depths greater than the first depth. The first rate is less than the second rate. See also... Figure 10E The figure shows a cross-sectional view of sample 1000e. The diameter of the hole 1008e within sample 1000e increases at a first rate with depth for depths less than a first threshold depth (not indicated), decreases at a second rate for depths greater than the first threshold depth but less than a second threshold depth, and decreases at a third rate for depths greater than the second threshold depth. The second rate is less than the third rate.

[0324] exist Figures 9A to 9E In each case, the (real) hole profile is depicted by a hyperbolic curve, while the estimated hole profile, derived from the (simulated) measured signal, is depicted by a dashed curve. The estimated hole profile corresponds to an estimate of the average hole profile (averaged over all holes in the array). However, since all holes are considered identical in the simulation, this distinction is irrelevant, except that boundary (i.e., edge) effects are reduced and a better estimate is obtained by simultaneously probing all holes in the array, rather than individual holes.

[0325] The simulated sound velocity within each aperture in the sample is practically independent of depth *s*, and furthermore, practically independent of aperture profile. That is, the simulated sound velocity depends essentially only on the material composition of the sample (which is the same for all samples). Local Brillouin frequency *f* B (s) is actually entirely composed of n eff (s) decision.

[0326] The local Brillouin frequency can be plotted as being correlated with the aperture diameter in a roughly linear manner. That is, f B (i) (s)~ai +b i ·d i (s), where the index i = 1, 2, ..., 5 labels the sample, and d i (s) is the diameter of the hole (in sample i) at depth s, and a i and b i (It is positive) is a constant. More specifically, the linear fitting algorithm is used to fit f with respect to the corresponding real hole contour. B (i) (s). In the real-world (i.e., non-simulated) implementation of method 300, the obtained local Brillouin frequencies can be linearly fitted with respect to the desired hole profile.

[0327] Figures 11A to 11C According to some embodiments, computer simulation results of depth profiling of three samples using method 300 are provided respectively. Figure 11A , Figure 11B and Figure 11C The samples correspond to Figure 9A , Figure 9B ,and Figure 9D Those. As detailed below, the significant differences in the estimated profiles between them are attributed to the use of different data analysis and fitting schemes.

[0328] More specifically, in order to obtain Figures 11A to 11C The estimated contour is obtained, and each element in the extracted signal is smoothed to identify its thermo-optical contribution. The thermo-optical contribution is then subtracted from the corresponding (unsmoothed) extracted signal, and the corresponding correlation of the Brillouin frequency with depth *s* is obtained. The obtained local Brillouin frequencies are then fitted with respect to the corresponding real hole contour using a third-order polynomial fitting algorithm. In a practical implementation of method 300, the obtained local Brillouin frequencies can be fitted with respect to the desired hole contour.

[0329] Used to obtain separately Figures 9A to 9E and Figures 11A to 11C Different data analysis and fitting schemes for the hole profiles produced results that were roughly similar in quality but still significantly different. Although Figures 9A to 9E The estimated hole profile is "noisy". Figures 11A to 11C The estimated hole profile is smooth, but by contrast, it exhibits a "systematic" error in the sense of overestimating or underestimating the diameter over an extended depth range. This suggests that a better estimate can be obtained. In particular, it is desirable to use machine learning or deep learning tools to generate a better estimate.

[0330] See Figure 12A and Figure 12B , Figure 12AThe illustration shows the penetration of a simulated x-polarized probe pulse into a simulated V-NAND stack 1200. More specifically, a cross-sectional view of the V-NAND stack 1200, including vias 1208, is shown. The V-NAND stack 1200 is a specific embodiment of sample 600. An intensity scale I (in arbitrary units) ranging from dark to light is also illustrated. The bottom of the scale corresponds to the minimum intensity, and the top of the scale corresponds to the maximum intensity, which is recorded inside the vias. Within the body of the V-NAND stack 1200, the intensity distribution of the probe pulse is seen to be maximum along the column of vias 1208 (extending parallel to the y-axis). That is, the x-polarized probe pulse penetrates into the body of the V-NAND stack between adjacent pairs of vias along the column, and substantially does not penetrate between adjacent pairs of vias along the row of vias 1208 (extending parallel to the x-axis). Figure 12B The illustration shows a simulated y-polarized probe pulse penetrating into a V-NAND stack 1208. Within the body of the V-NAND stack 1208, the intensity distribution of the probe pulse is observed to be maximum along the rows. That is, the y-polarized probe pulse penetrates into the body of the V-NAND stack between adjacent pairs of vias along the rows, and substantially does not penetrate between adjacent pairs of vias along the columns.

[0331] Figure 13 The extracted signal obtained by computer simulation of an implementation of FinFET method 300 is depicted according to some embodiments. FinFET is constructed from a sample 700 ( Figures 7A to 7E The fin profile of fin 708 (described in the text) is characterized. More specifically, the (normalized) extracted signal E obtained when a series of probe pulses are scattered away from the acoustic pulse at different depths within the fin is depicted. (E = ΔI / I, as described above in the text) Figure 8 (As explained in the description above.) Figures 7A to 7E The description states that the acoustic pulse is generated by a consistently prepared pump pulse. The horizontal axis measures the time t from the generation of the acoustic pulse, or the scattering time of the same object. Specifically, the larger t is, the greater the scattering depth. At time t = 0, the acoustic pulse begins to propagate from the absorption layer (i.e., the top layer of the fin (e.g., absorption layer 712)) into the fin. The maximum scattering depth corresponds to t = t_0. 基底 (At this point, the acoustic pulse reaches the substrate portion of the FinFET, for example, substrate portion 705). A single Brillouin oscillation is observed.

[0332] Figures 14A to 14EAccording to some embodiments, computer simulation results of depth profiling of five samples using method 300 are provided respectively. Each of the five samples includes a plurality of parallel and consistent fins (e.g., fin 708) disposed on a substrate portion (e.g., substrate portion 704). The samples differ from each other in terms of the lateral cross-sectional profile of the fins. Each of the five modeled samples is similar to sample 700 and therefore may correspond to a possible design of a fin field-effect transistor (FinFET), or to a specific possible distortion attributable to manufacturing defects in such a design. The vertical axis parameter is such as the depth s measured from the top of the fin. (The top layer of the fin constitutes an absorption layer, as described above in...) Figures 7A to 7E As described in the description. ) The horizontal axis x parameterizes the width of the fin. Figures 14A to 14E Each of the fins is depicted as only half of its fin, implicitly understood to exhibit mirror symmetry about the ys plane (where the y-axis points outward from the page). More specifically, the yx plane divides each fin into two coherent longitudinal portions.

[0333] In each of the accompanying figures, the (real) fin profile is depicted by a double-line curve, while the estimated fin profile, derived from the (simulated) measured signal, is depicted by a dashed curve. The estimated fin profile corresponds to the estimate of the average fin profile. However, since all fins are considered identical in the simulation, this distinction is irrelevant, except that by simultaneously probing all fins in the array, rather than a single fin, boundary effects are reduced and a better estimate is obtained.

[0334] Linear regression algorithms are used to estimate the average fin profile based on the extracted signal. More specifically, temporal linear regression algorithms are used to correlate scattering time with the fin width (at the depth s where scattering occurs). The scattering time t is determined via s = v 声 ·t, is directly related to the scattering depth s, thus allowing the (average) width of the fin to be (estimated) in relation to the scattering depth s.

[0335] In each of the simulations, the pump pulse and probe pulse are linearly polarized along the longitudinal dimension of the fin (i.e., parallel to the y-axis). For example, in Figure 7E As explained in the description, this choice of pump pulse polarization minimizes the penetration of the pump pulse into the sidewalls of the fin, and thus helps to generate a uniform acoustic pulse within each of the fins. Additionally, as can be seen in... Figure 15A and Figure 15B As can be seen, this choice of probe pulse polarization maximizes the interaction between the probe pulse and the acoustic pulse.

[0336] Figure 15AThe illustration shows a y-polarized probe pulse penetrating into multiple parallel fins 1508. Fins 1508 protrude from a base portion 1504. Fin 1508 is a specific embodiment of fin 708. Base portion 1504 is a specific embodiment of base portion 704. The vertical axis s is parameterized as the depth measured from the top of the fin. The horizontal axis x extends perpendicular to the elongation dimension of fin 1508. More precisely, Figure 15A The diagram illustrates a cross-sectional view of fin 1508 and base portion 1504, overlaid with the intensity distribution of the probe pulse, expressed in grayscale. The intensity scale I (in arbitrary units) quantizing the intensity from dark to light is also illustrated. The bottom of the scale corresponds to the minimum intensity, and the top of the scale corresponds to the maximum intensity. The probe pulse can be clearly seen penetrating into fin 1508.

[0337] In comparison, Figure 15B The illustration shows a comparison of x-polarized probe pulses not penetrating into the fin: the probe pulses essentially do not penetrate into the fin.

[0338] As used herein, the terms “lateral extension” and “maximum lateral extension” are used interchangeably with respect to lateral structural features or composite lateral structural features.

[0339] As used herein, the terms “depth profiling” and “3D probing” may be used interchangeably according to some embodiments.

[0340] It will be understood that, for clarity, certain features of this disclosure described in the context of different embodiments may also be provided in combination in a single embodiment. Conversely, for simplicity, various features of this disclosure described in the context of a single embodiment may also be provided separately or in any suitable sub-combination or appropriately provided in any other described embodiment of this disclosure. Unless expressly stated otherwise, features described in the context of an embodiment are not considered essential features of those embodiments.

[0341] Although operations in the disclosed methods may be described in a specific order according to some embodiments, the methods of this disclosure may include some or all of the described operations performed in a different order. The methods of this disclosure may include some or all of the described operations. Unless expressly stated otherwise, specific operations in the disclosed methods are not considered essential operations of those methods.

[0342] Although this disclosure has been described in conjunction with specific embodiments thereof, it will be apparent that many alternatives, modifications, and variations are possible and will be obvious to those skilled in the art. Therefore, this disclosure encompasses all such alternatives, modifications, and variations that fall within the scope of the appended claims. It will be understood that this disclosure is not necessarily limited in its application to the details of the construction and arrangement of the components and / or methods set forth herein. Other embodiments may be practiced, and embodiments may be implemented in various ways.

[0343] The wording and terminology used herein are for descriptive purposes and should not be construed as restrictive. Any reference or identification in this application should not be interpreted as an admission that such reference is available as prior art in this disclosure. Section headings are used herein to facilitate understanding of the specification and should not be construed as necessary limitations.

Claims

1. A method for depth profiling of a sample, the method comprising the following operations: A sample is provided, the sample containing a target region, the target region containing lateral structural features; Multiple measured signals can be obtained by performing the following sub-operations multiple times: An optical pump pulse is projected onto the sample to generate an acoustic pulse that propagates within the target region of the sample, wherein the wavelength of the pump pulse is greater than at least about twice the lateral range of the lateral structural feature along at least one lateral direction. An optical probe pulse is projected onto the sample, such that the probe pulse undergoes Brillouin scattering as it exits the acoustic pulse within the target region; as well as The scattered component of the probe pulse is detected to obtain the measured signal; In each implementation, the corresponding detection pulse is scattered away from the acoustic pulse at a corresponding depth within the target area, thereby enabling the detection of the target area at multiple depths; as well as Analyze the multiple measured signals to obtain the depth correlation of at least one parameter characterizing the lateral structural features.

2. The method of claim 1, wherein the propagation direction of the acoustic pulse within the target region is parallel to the longitudinal dimension of the target region, which parameterizes the depth within the sample, and wherein the probe pulse is configured such that the absorption length of the probe pulse within the target region is greater than the range of the target region along the longitudinal dimension.

3. The method of claim 1, wherein the lateral structural feature is manifested as a change in refractive index and / or a change in sound velocity in the target region along at least one lateral direction.

4. The method of claim 3, wherein the change in refractive index and / or the change in sound velocity is attributed to one or more of the following design-related changes in the target region along the at least one lateral direction: geometry, material composition, medium, mass density, density of embedded components and / or pores, and spatial arrangement of embedded components and / or pores; and The at least one parameter characterizing the transverse structural features includes one or more parameters characterizing the geometry, material composition, medium, mass density, density of embedded components and / or pores, and / or spatial arrangement of embedded components and / or pores.

5. The method of claim 1, wherein in the analysis of the plurality of measured signals, a predetermined desired depth correlation of the at least one parameter characterizing the lateral structural features is taken into account.

6. The method of claim 1, wherein, optionally, in the analysis of the plurality of measured signals, a data fitting tool derived from machine learning and / or deep learning techniques is used to obtain the deep correlation of the at least one parameter.

7. The method of claim 1, wherein the analysis of the plurality of measured signals comprises obtaining the time correlation of the frequency and / or the time correlation of the amplitude of the Brillouin oscillation characterizing the scattering component of the probe pulse, and based thereon obtaining the depth correlation of the at least one parameter characterizing the lateral structural feature.

8. The method of claim 1, wherein the analysis of the plurality of measured signals comprises removing the thermo-optical contribution to the plurality of measured signals.

9. The method of claim 1, wherein the frequency of the pump pulse and / or the frequency of the probe pulse maximizes or substantially maximizes the intensity of the scattering component of the probe pulse; and / or The pump pulse and / or the probe pulse are polarized to maximize or substantially maximize the intensity of the scattering component of the probe pulse.

10. The method of claim 1, wherein each of the pump pulses is configured to induce mechanical strain in one or more transverse absorption layers of the sample, thereby generating a corresponding acoustic pulse, and wherein the one or more absorption layers are perpendicular to the longitudinal dimension of the target region parameterizing the depth within the sample.

11. The method of claim 10, wherein the one or more absorption layers are silicon-based, the duration of each of the pump pulse and the probe pulse is less than about 10 psec, and the width of the acoustic pulse is less than about 300 nm.

12. The method of claim 10, wherein the target region comprises the one or more absorption layers.

13. The method of claim 10, wherein the absorption layer is positioned adjacent to the lateral outer surface of the sample on which the pump pulse and / or the probe pulse is projected, such that each of the acoustic pulses propagates away from the lateral outer surface of the sample; or The absorption layer is positioned within the sample such that each of the acoustic pulses propagates away from the absorption layer toward the lateral outer surface on which the pump pulse and / or the probe pulse is projected.

14. The method of claim 1, wherein the target region comprises a plurality of said lateral structural features, the plurality of said lateral structural features defining a composite lateral structural feature, wherein the wavelength of the pump pulse and the wavelength of the probe pulse are configured to allow simultaneous detection of the composite lateral structural feature, and wherein, in the operation of analyzing the plurality of measured signals, the obtained depth correlation of the at least one parameter characterizing the lateral structural feature is the average depth correlation of the plurality of said composite lateral structural features.

15. The method of claim 14, wherein the sample is a fin field-effect transistor (FinFET), wherein the target region comprises a plurality of fins arranged parallel to each other to form the composite lateral structural feature, wherein the at least one parameter characterizing the lateral structural feature includes a parameter corresponding to the average width of the fins, and optionally, wherein the pump pulse and the probe pulse are linearly polarized parallel or substantially parallel to the elongation dimension of the fins.

16. The method of claim 14, wherein the sample is a vertical NAND stack, wherein the target region includes a plurality of holes projected into the target region in a longitudinal dimension parallel to the target region, the longitudinal dimension parameterizing the depth within the vertical NAND stack, wherein the holes are configured to form the composite lateral structure feature, and wherein the at least one parameter characterizing the composite lateral structure feature includes a parameter corresponding to the average diameter or average area of ​​the holes.

17. The method of claim 16, wherein the apertures are arranged in a two-dimensional rectangular array, the probe pulses are linearly polarized along a lateral direction parallel to a first direction defined by rows of the two-dimensional rectangular array or a second direction defined by columns of the two-dimensional rectangular array, thereby increasing measurement sensitivity along the second direction or the first direction, respectively.

18. The method of claim 1, wherein the probe pulse is characterized by a first probe wavelength and / or a first probe polarization, and / or wherein the pump pulse is characterized by a first pump wavelength and / or a first pump polarization; The method further comprises, prior to analyzing the plurality of measured signals, repeatedly obtaining the plurality of measured signals by: (i) a second probe pulse characterized by a second probe wavelength and / or a second polarization, and / or (ii) a second pump pulse characterized by a second pump wavelength and / or a second pump polarization, thereby obtaining a second plurality of measured signals; and In the operation of analyzing the plurality of measured signals, the plurality of measured signals are analyzed together with at least the second plurality of measured signals.

19. A computerized system for depth profiling of a sample, the system comprising an optical setup and a measurement data analysis module, wherein the optical setup is configured to acquire a plurality of measured signals from a target region of the sample by repeating the following operations: An optical pump pulse is projected onto the sample to generate an acoustic pulse that propagates within the target region, wherein the wavelength of the pump pulse is at least about twice the lateral extent of a lateral structural feature in the target region along at least one lateral direction. An optical probe pulse is projected onto the sample, such that the probe pulse undergoes Brillouin scattering as it exits the acoustic pulse within the target region; as well as The scattering component of the probe pulse is detected, thereby obtaining the corresponding measured signal from the plurality of measured signals; In each repetition, the detection pulse is scattered away from the acoustic pulse at a corresponding depth within the target area; and The measurement data analysis module is configured to analyze the plurality of measured signals or a plurality of signals derived from the plurality of measured signals to obtain the depth correlation of at least one parameter characterizing the lateral structural features.

20. A non-transitory computer-readable storage medium storing instructions that cause a sample analysis system to perform: Multiple measured signals of a sample containing a target region, which includes lateral structural features, are obtained by performing the following sub-operations multiple times: An optical pump pulse is projected onto the sample to generate an acoustic pulse that propagates within the target region of the sample, wherein the wavelength of the pump pulse is greater than at least about twice the lateral range of the lateral structural feature along at least one lateral direction. An optical probe pulse is projected onto the sample, such that the probe pulse undergoes Brillouin scattering as it exits the acoustic pulse within the target region; as well as The scattered component of the probe pulse is detected to obtain the measured signal; In each implementation, the corresponding detection pulse is scattered away from the acoustic pulse at a corresponding depth within the target area, thereby enabling the detection of the target area at multiple depths; as well as Analyze the multiple measured signals to obtain the depth correlation of at least one parameter characterizing the lateral structural features.

Images